GARWOOD, Circuit Judge:
The defendant-appellant, Freightliner Corporation, appeals from a judgment on a jury verdict in favor of the plaintiffs-appellees, the Sodens, in this Texas law diversity action based on products liability. The plaintiffs-appellees are the widow and two children of Willard C. Soden, who was killed in a post-collision fuel fire while driving a truck designed and manufactured by Freightliner. The Sodens claimed that the design of the Freightliner fuel system was unreasonably dangerous and was the producing cause of the post-collision fuel fire which killed Willard C. Soden. On appeal, Freightliner argues that the district court committed reversible error with respect to three evidentiary rulings made during the trial, which (1) excluded statistical and opinion evidence offered by its expert, Mr. Hutton, (2) admitted with a limiting instruction evidence of other lawsuits against Freightliner involving accidents allegedly caused by defects in its fuel system, and (3) admitted opinion testimony by Robert Lasr ere, a lay witness, respecting the dangerousness of the step brackets mounted on the fuel tank of Soden’s truck. Finding no reversible error, we affirm the judgment below.
I.
CONTEXT FACTS
On April 18, 1978, Willard C. Soden was driving a tractor trailer carrying a load of crude oil. The tractor was designed and manufactured by the defendant-appellant Freightliner. The trailer was made by another manufacturer. Soden’s vehicle collided with a Ford Mustang at an intersection and rolled over onto its right side. Soden was not at fault in the accident. He immediately climbed up through the left door onto the outside of his vehicle’s cab, but while jumping from it he became engulfed in fire. He died the following day from burns he received in that fire.
The plaintiffs, Soden’s widow and minor children, alleged that the defective design of the fuel system in the Freightliner truck was a producing cause of the fire which resulted in Soden’s death. The thrust of their argument was that the cylindrical fuel tanks, mounted on the sides of the truck under the cab doors, were improperly protected and subject to breaking up and causing a fire in the event of a collision. In particular, they also argued that the brackets securing the steps to these fuel tanks had pointed ends which, in the event of a rollover, could puncture the fuel tanks. The resulting hole or holes could release diesel fuel near engine components hot enough to ignite the fuel, causing a fire in the engine-cab area. They sought $1,400,-000 in compensatory damages and $10,000,-000 in exemplary damages allegedly due because of Freightliner’s “gross negligence” in failing to correct the design of the fuel system after having notice of its defects.
Freightliner took the position that the fuel tanks were located in the only position feasible for trucks of such design and that the step bracket did not, or could not have, punctured the fuel tank. It contended that the initial source of the fire was the crude oil from the trailer Soden was pulling. It also maintained that it had not been grossly negligent.
The jury held in the appellees’ favor and awarded them a total of $885,300 in compensatory damages, but nothing in exemplary damages. Freightliner argues on appeal that the trial judge committed prejudicial error in the three evidentiary rulings discussed below.
II.
CLAIMS ON APPEAL
A. EXCLUSION OF FREIGHTLINER’S STATISTICAL AND OPINION EVIDENCE.
Freightliner, on direct examination, tried to elicit from Mr. Hutton, their expert and [501]*501representative at trial, certain statistics on accidents involving fires in Freightliner trucks to rebut statistics offered by the Sodens on the frequency of fuel system fires in large trucks. These statistics and Hutton’s calculations based upon them were also summarized on two charts which Freightliner attempted to introduce into evidence: one chart broke down the number of accidents involving fires into subcategories of (1) electrical system fires, (2) fuel system fires, (3) other classifications of fire, and (4) post-crash fuel system fires; and a second chart depicted Hutton’s calculation that post-crash fuel system fires occurred only once every 409,785,800 truck-driving miles. Hutton was presumably prepared also to offer his opinion on the significance of these statistics.
Hutton asserted that the statistics themselves were based on data collected over a three-year period by a sister company, Consolidated Freightways, from its fleet of roughly 2,200 Freightliner trucks and purported to show that only three out of 3,014, or one-tenth of one percent of the accidents which occurred, were accompanied by a post-crash fuel system fire. Hutton’s source for this information was Ms. Vorman, the assistant to the director of safety at Consolidated Freightways, who, according to what Vorman told Hutton, had compiled it from 3,000 accident reports made over the three-year period to the Bureau of Motor Carriers Safety on the Bureau’s forms. Vorman, who did not testify, provided Hutton this information in a two-page letter (which was shown to the district court but not introduced into evidence) and in a telephone conversation.
Freightliner’s counsel attempted to introduce the statistics and the two charts into evidence through Hutton in the course of the following examination:
“Q And have you had occasion to look in to and research the [occurrence] of accidents with these trucks?
“A I have attempted to do that from every available source that you can think of. And for the most part, there is much slack in available data.
“Between moving vans and cars, and it is really hard to say, you know, heavy duty combination vehicles such as the one Mr. Soden was driving, it is hard to find good data on that.
“Q And from the findings that you have made, do you find that the fuel system usually is or is not involved in the various wrecks that these trucks get involved in?
“A I think we have heard or seen all kinds of numbers batted around back and forth here. And the study I have the most confidence in, I have found it is about one-tenth of one percent of the accidents of the Freightliner trucks are apt to result in a post-crash fire that injures the fuel system.
“Q And have you accumulated some statistics on that particular point?
“A Yes, I have.
“Q And has a chart been prepared under your supervision?
“A Yes.
“Q Based on those statistics — ”1 At this point the Sodens’ counsel interposed an objection based on a motion in limine to exclude Freightliner’s two charts.2
The district court, after extensive discussion with counsel and the voir dire of Hutton by both counsel, granted the motion to exclude the evidence. The district court’s central concern was that Freightliner had not offered an adequate predicate for the statistics to which Hutton was testifying, which he had summarized on the charts, and on which he was attempting to base an opinion:
[502]*502“And I am going to grant the plaintiffs motion in limine because it appears to me that there has been a failure to [comply] with a rule as well as an attempt to give a summary of testimony of a witness who has purported to have made the examination.
“But none of the supporting data nor is this examining data witness ... available to the Court. This witness Gene [sic] Vorman is not available.
“Now, this is without prejudice to the defendant bringing in the witness Gene [sic] Vorman. And this data and affording plaintiff’s counsel an opportunity to review the supporting data and a chance to be heard in the absence of the jury if it is to be requested.”
In the discussion that followed, Freight-liner’s counsel asked the court, “You are excluding the charts? Are you excluding the evidence concerning it?” And the court responded, “I am [excluding the evidence] concerning everything until you come in with a better predicate and then I may after I have heard that, I may give it under a limiting instruction even then.” The district court then instructed the jury to disregard the statistical evidence given by Hutton immediately prior to the objection.
The parties dispute whether the district court’s ruling excluded Hutton’s opinion based upon the statistics, as well as the figures themselves and the charts on which they were depicted.3 We assume, however, from the court’s response to the question by Freightliner’s counsel that Hutton’s opinion was also held inadmissible to the extent that it was dependent on the statistics. Even so, the district court’s exclusion of this evidence was within its discretion. We examine first the excluded opinion testimony.
Rule 703 of the Federal Rules of Evidence states that:
“The facts or data in the particular case upon which an expert bases an opinion or inference may be those perceived by or made known to him at or before the hearing. If of a type reasonably relied upon by experts in the particular field in forming opinions or inferences upon the subject, the facts or data need not be admissible in evidence.”
This Rule broadened considerably the permissible basis for expert opinions, which had been limited at common law to
“(1) information obtained by the expert’s personal knowledge; (2) the hypothetical question, which assumes facts reasonably supported by the evidence; and (3) in some jurisdictions, testimony previously elicited during the trial, with the expert instructed to assume the truth of that evidence and to base his conclusions thereon.” Zenith Radio Corp. v. Matsushita Electric Industrial Co., 505 F.Supp. 1313, 1322 (E.D.Pa.1980).
The Rule was intended, as described in the Notes of the Advisory Committee on Proposed Rules, to “bring the judicial practice into line with the practice of the experts themselves when not in court.” Thus, under Rule 703 an expert can discuss as the basis for an opinion facts or data which are otherwise inadmissible hearsay, “[i]f of a type reasonably relied upon by experts in the particular field in forming opinions or inferences upon the subject.”4 A trial [503]*503court’s inquiry into whether this standard is satisfied must be made on a case-by-case basis and should focus on the reliability of the opinion and- its foundation rather than merely on the fact that it was based, technically speaking, upon hearsay. See Bauman v. Centex Corporation, supra, 611 F.2d at 1120 n. 5. A district court’s determination in this regard will be reversed only if it constitutes an abuse of discretion. Id. at 1120; accord, Seese v. Volkswagenwerk A.G., 648 F.2d 833, 844 (3d Cir.), cert. denied, 454 U.S. 867, 102 S.Ct. 330, 70 L.Ed.2d 168 (1981).
Assessing Hutton’s statistics, as presented to the court below, we conclude that they were not shown to be of a type reasonably relied upon by experts in his field, and thus the district court did not abuse its discretion in excluding any opinion dependent on them. First, the statistics on which Hutton relied were prepared strictly in anticipation of litigation and were based on information received from a sister company. They were not part of a published study and were not even made available to the appellees until Freightliner attempted to introduce them.5 These facts standing alone do not suffice to impugn the reliability of the statistics, but do properly signal a trial judge to make a critical review of their bases.
Second, the only memorialization of what Hutton was relying on, namely, Yorman’s summary of her analysis of certain Consolidated Freightways records made at Hutton’s request for this case, was extremely informal — a two-page letter to Hutton, not introduced into evidence, and a follow-up telephone conversation with Hutton.6 The telephone conversation occurred because of a misunderstanding with respect to Hutton’s original request. No record of this conversation existed except as a notation by Hutton on the original letter. Third, as Hutton seemed to acknowledge, a review of each of the 3,000 reports was a major task, yet nowhere did he explain the method by which Vorman accomplished it: whether the summary of the data from these reports was made on a regular basis (which it ap[504]*504parently was not); whether Yorman herself read and analyzed each of the 3,000 reports or supervised employees in that task; and whether and in what form the Bureau of Motor Carriers Safety accident reports required information about post-crash fuel system fires.
Finally, in addition to these factors which suggested weaknesses in the reliability of Hutton’s statistics, the Sodens raised a more fundamental question which placed serious doubt upon their relevance. Hutton’s two-page letter from Vorman stated that “ ‘[d]uring the period of ’75 through ’78, we found fifteen road accidents which were caused by some type of fuel fire.’ ” (Emphasis added.) The appellees argued that this information failed to include accidents not caused by fire, but in which post-crash fires resulted. Hutton responded that he got this information in the follow-up telephone conversation.7
The appellees, however, offered the annual summary of these Bureau of Motor Carriers Safety reports for 1977, which apparently included information only on the causes of accidents, as proof that these reports did not include information with respect to post-crash fuel system fires.8 This possibili[505]*505ty cast doubt on how Vorman could have derived the necessary information, and Hutton was without question relying on her alone for the statistics which formed the basis of his opinion.
This doubt involved not merely the weight Hutton’s statistics and opinion should have been accorded, but whether they had any meaningful relevance at all to the frequency of post-crash fuel fires in Consolidated Freightways’ fleet of Freight-liner trucks. Were the Sodens, in fact, correct in their assessment of Vorman’s summary, Hutton’s opinion based upon it would have been misleading and largely irrelevant. Though courts have afforded experts a wide latitude in picking and choosing the sources on which to base opinions, Rule 708 nonetheless requires courts to examine the reliability of those sources. See Zenith Radio Corp. v. Matsushita Electric Industrial Co., supra, at 1323-24; 3 J. Weinstein & M. Burger, Weinstein’s Evidence, ¶ 703[03] at 703-16 (1982); 3 D. Louisell & B. Mueller, Federal Evidence, § 389 at 661-62 (1979). Thus, the district court quite properly used its discretion to exclude Hutton’s opinion until a better predicate for his statistics could be presented.9
Even with this opportunity, however, Freightliner made no further attempt to satisfy the district court’s concerns, even though the presentation of evidence did not conclude until four days later and Hutton testified again on the last day. Nor did Freightliner request a continuance or indicate to the court that any practical difficulty stood in the way of its establishing a better predicate, either because Vorman was unavailable to testify or because the records could not be reproduced. Given the concerns which were brought to the district court’s attention and which went to the relevance and reliability of Hutton’s statistics, exclusion of his opinion testimony [506]*506based upon them clearly was not an abuse of discretion.
Likewise, the district court’s exclusion of the two charts Hutton had prepared and its instruction to the jury to disregard Hutton’s mention of the statistics before the objection were also proper. Hutton’s testimony respecting the statistics was inadmissible as direct evidence of the frequency of post-crash fuel fires in Freight-liner trucks because it was hearsay, being based on a letter and a telephone conversation. See United States v. Brown, 548 F.2d 1194, 1206 n. 22 (5th Cir.1977); cf. United States v. 478.34 Acres of Land, 578 F.2d 156, 159 (6th Cir.1978). The charts were also correctly excluded because, as summaries of Hutton’s statistics and calculations thereon, they could not be admitted unless the underlying information was also admissible. Gordon v. United States, 438 F.2d 858, 876-77 (5th Cir.), cert. denied, 404 U.S. 828, 92 S.Ct. 139, 30 L.Ed.2d 56 (1971); United States v. Kim, 595 F.2d 755, 764 (D.C.Cir.1979). See also United States v. Foshee, 606 F.2d 111, 113 (5th Cir.1979), cert. denied, 444 U.S. 1082, 100 S.Ct. 1036, 62 L.Ed.2d 766 (1980).10
Freightliner argues in opposition to the exclusion of the opinion evidence that courts have recognized that experts often rely on third parties for facts and data in expressing opinions under Rule 703. It argues that the doubts raised as to the reliability and relevance of Hutton’s statistics were mistaken and that in any event cross-examination, not exclusion, was the proper means of handling any weaknesses in Hutton’s opinion. We reject these arguments.
Experts may under certain circumstances rely on the information received from third parties, but in this case reliance on the third party was not the only significant source of concern.11 The most troubling aspect of the evidence lay in whether the reports submitted to the Bureau of Motor Carriers Safety actually contained regularly reported information about post-crash fuel fires. Freightliner argues that the doubts raised in this regard were mistaken. It asserts on appeal that the “Bureau of Motor Carriers Safety define[s] a ‘reportable accident’ as one resulting in a fatality, injury, or certain property damage.” Since post-crash fires result in property damage, they are also included in the accident reports.
This argument, although it may be valid, was not made to the district court, which had before it the Vorman letter which stated that it was listing only accidents which were caused by fire. Nor did Freightliner explain the particulars of how Vorman assembled the information given to Hutton, which might have described the feature of the accident reports which made the number of post-crash fires available from them. The court did have before it an indication that post-crash fuel fires were not included in the Bureau reports, which was never squarely rebutted by Freightliner. Given the evidence concerning the statistics which was presented, the district court properly exercised its discretion in requiring that Freightliner offer a better predicate for the statistics before it would admit Hutton’s opinion dependent upon them. Vorman was not available for cross-examination. Hutton could not be cross-examined on the Bureau of Motor Carriers reports because they were not available and he had not examined them. Moreover, as noted, the concerns respecting these statistics extended to their relevance, not simply their weight. Accordingly, we also reject Freightliner’s argument that cross-examination was available to bring out any weak[507]*507nesses in Hutton’s testimony, and that his opinion, therefore, should not have been excluded.
B. EVIDENCE OF OTHER LAWSUITS AGAINST FREIGHTLINER.
Over Freightliner’s objection the district court admitted, with a limiting instruction, the appellees’ evidence that five suits and at least three informal complaints had been filed against Freightliner, which included allegations of “injuries to users and bystanders of [Freightliner] trucks from post-collision diesel fuel fires.” The following is the sum of that testimony, elicited from Hutton on cross-examination by the appellees’ counsel:
“Q Now, sir, isn’t it true that prior to the manufacture and sale of the truck that is involved in this case, your company had actual knowledge of injuries to users and bystanders of these trucks from post-collision diesel fuel fires, you actually knew that this was occurring?
“A We had some knowledge of that.
“Q And isn’t it also true that your knowledge included the fact that your company had been sued in litigation where these allegations had been made?
“A I am aware of a number of litigation cases; yes.
“Q In fact, in at least five former litigation cases that preceded this one as well as some three to five informal complaints that had not yet come to litigation?
“A Somewhere in that neighborhood, I don’t know the exact numbers.”
At this point, the district court gave a limiting instruction, which stated that the existence of the allegations in the suits was no evidence of their truth, but was only evidence of Freightliner’s notice of claims of defective design.12
Freightliner does not object to the first question and answer quoted above or to the mention of the informal complaints, but only to the mention of the lawsuits which had been filed.13 It argues on appeal (1) “that the mere existence of the suits is not relevant to notice,” and (2) that if it is, “its [508]*508potential for unfair prejudice so outweighs its weak probative value that the admission of the evidence was an abuse of discretion.”
We hold that the district court’s decision to admit the evidence of the prior litigation was not reversible error. The So-dens offered this evidence to show that Freightliner had notice of the defective design of its fuel system, which was an element of their claim for exemplary damages. In a products liability case based on defective design, evidence of the defendant’s knowledge of other accidents may be relevant to its notice or knowledge of the defect, if the accidents occurred under substantially similar circumstances. Prashker v. Beech Aircraft Corporation, 258 F.2d 602, 608 (3d Cir.) cert. denied, 358 U.S. 910, 79 S.Ct. 236, 3 L.Ed.2d 230 (1958); Ramos v. Liberty Mutual Insurance Co., 615 F.2d 334, 338 (5th Cir.1980), cert. denied sub nom. Rucker Co. v. Shell Oil Co., 449 U.S. 1112, 101 S.Ct. 921, 66 L.Ed.2d 840 (1981). Such evidence is relevant because it makes more probable “a fact that is of consequence to the determination of the action” — in this case Freightliner’s knowledge of the defective design of its fuel system, which was a required element in the appellees’ claim for exemplary damages. See Burk Royalty Co. v. Walls, 616 S.W.2d 911 (Tex.1981); Maxey v. Freightliner Corp., 665 F.2d 1367 (5th Cir.1982).14
Some decisions have held evidence of pri- or litigation, introduced by means of a complaint, relevant on the question of notice if the complaint contains allegations of a design defect causing an accident under circumstances substantially similar to those in the case before the Court. Gulf States Utilities Co. v. Ecodyne Corp., 635 F.2d 517, 519 (5th Cir.1981); Julander v. Ford Motor Co., 488 F.2d 839, 846 (10th Cir.1973) (excluding the complaints as notice of defective design, however, because they were not prior to the plaintiff’s accident). See also Gardner v. Q.H.S., Inc., 448 F.2d 238, 244 (4th Cir.1971) (“letters of complaints [nonlitigation] should have been admitted where they described identical or similar experiences”; “specific letters may be withheld from the jury by the district judge, in the exercise of his discretion” on Rule 403 grounds); New York Life Insurance Co. v. Seighman, 140 F.2d 930, 932-33 (6th Cir. 1944).
Evidence of allegations of defective design arising out of other accidents is obviously less probative on the question of notice than evidence which shows the particular circumstances of the other accidents, how the design defect was involved in them, and how the company became aware of them. Allegations in a complaint may be unsubstantiated and because of their hearsay character they are not subject to cross-examination. Certainly there may be potential dangers in this area, and caution should be exercised when dealing with mere allegations, even to prove notice. Nonetheless, allegations alone may, depending on the circumstances, have some probative value as evidence of notice, and the trial judge generally has broad discretion in assessing their worth and possible prejudice.
Turning to the case at bar, the evidence of the other lawsuits involved allegations of post-crash fuel fires in Freightliner trucks caused by the defective design of the fuel system, which were the same allegations being made by the plaintiffs. Thus, the substantial similarity requirement was satisfied with respect to the allegations made in the suits. In addition, the weaknesses otherwise inherent in allegations as a basis for notice were mitigated by several factors. First, Freightliner did not contest, and it is apparent that the parties proceeded at trial on the assumption, that the underlying accidents actually occurred, and that they involved post-crash fuel fires. There is no suggestion Freightliner believed otherwise at any relevant time. Freightliner, and often Hutton himself, investigated [509]*509these accidents on a regular basis. Second, the district court gave Freightliner the opportunity to rebut its conclusion that the allegations were “sufficiently similar” to the facts of the case before it because they involved attacks “on the defective design of the fuel system”:
“THE COURT: Mr. Hebdon, if they are not of the same type, do you want to go in to it to show they are not the same[?] I will hear you and I will entertain the motion to strike.
“MR. HEBDON: I understand, sir.
“THE COURT: The only way I can do it is [to proceed] with the whole interrogation. If you think that is necessary, I will do it.”
Freightliner, however, made no attempt after this offer to analyze or show that the accidents involved in the suits were not similar.15 Finally, it was arguably to Freightliner’s advantage that facts of the specific accidents (the actual occurrences of which were not really disputed, as we have noted) not be brought out in detail to show the jury the degree of similarity of the other accidents by direct evidence. Thus, although the use of unsupported allegations in prior lawsuits to prove notice may not have been of the strongest probative value, neither was the lack of support particularly prejudicial.
In this regard we note that although Freightliner argues on appeal that this evidence was unduly prejudicial under Rule 403, it did not raise that issue at trial nor seek a ruling on it.16 Thus, we question whether this objection has been properly preserved for our review. See Falcon v. General Telephone Company of the Southwest, 626 F.2d 369, 382 (5th Cir.1980), vacated on other grounds, 450 U.S. 1036, 101 S.Ct. 1752, 68 L.Ed.2d 234 (1981). We conclude that the admission of the evidence did not substantially and unfairly prejudice Freightliner’s rights to a fair trial. We note at the outset that we are at a distinct disadvantage in weighing probative value against the danger of unfair prejudice for the first time on appeal. The plaintiffs and the trial court were not given an opportunity to address this contention, or to attempt to mitigate the complained of effects, at trial. Moreover, in this case the questions complained of were followed by a most thorough and proper limiting instruction, which we have set out, emphasizing certain particularly apt portions, in note 12, supra. This instruction properly made clear the hearsay, unsworn, and self-serving nature of the other complaints, and specified exactly what consideration the allegations in the other lawsuits were and were not to be given, and mitigated the danger of any unfair prejudice. See United States v. Moore, 522 F.2d 1068, 1079 (9th Cir.1975), cert. denied, 423 U.S. 1049, 96 S.Ct. 775, 46 L.Ed.2d 637 (1976) (“the judge’s careful limiting instructions undoubtedly minimized, if they did not eliminate, any unfair prejudice that might have accrued”); United States v. Catalano, 491 F.2d 268, 274 (2d Cir.), cert. denied, 419 U.S. 825, 95 S.Ct. 42, 42 L.Ed.2d 48 (1974).
Further, the mention of the lawsuits to which Freightliner objects on appeal consisted of two questions asked over the course of a two-week trial which generated over 1,000 pages of transcript. Their existence was not unfairly emphasized at the trial. And, the allegations in the other suits were somewhat probative on the issue [510]*510of notice given the fact that the accidents on which they were based admittedly included post-crash fuel fires involving Freightliner trucks.
Freightliner argues that the plaintiffs did not need this evidence to prove that it had some knowledge of post-crash fuel fires because Freightliner had admitted this much. Thus, the value of the evidence was weakened for the purposes of balancing under Rule 403. See 22 Wright & Graham, Federal Practice & Procedure § 5214 at 268-69 (1978). However, evidence of specific allegations of fuel system defects is clearly more probative on the question of knowledge than Freightliner’s admission that it had “some knowledge” of injuries from post-collision fuel fires generally.
On the degree of prejudice, Freightliner relies on Yellow Bayou Plantation, Inc. v. Shell Chemical, Inc., 491 F.2d 1239 (5th Cir.1974), in which the plaintiff sued a herbicide maker for damages caused by the ineffectiveness of its products. On appeal the Court upheld the exclusion of “a list of lawsuits over, and complaints to Shell about, Planavin [the herbicide], offered to show that Shell knew or should have known that Planavin was ineffective.” Id. at 1242-43. The Court concluded:
“The most that these items could have indicated was that absent third parties had made this claim to or against Shell from time to time. To exclude evidence of such faint probative value and high potential for unfair prejudice was well within the trial court’s discretion.” Id.
Here, as in Yellow Bayou, the discretion of the trial court, now reflected in Rule 403, points to affirmance of the trial court’s ruling, not reversal. We do not suggest that the trial court lacked discretion under Rule 403 to exclude the evidence. Rather, we hold that its admission was not reversible error.
Finally, it is certainly important to note that Freightliner prevailed on the issue of exemplary damages. The only issue to which the complained of evidence was relevant, as tending to show Freightliner’s knowledge, was resolved by the jury in Freightliner’s favor. See Ersler v. T.F. Schneider Corp., 188 F.2d 1022, 1023 (D.C. Cir.1951); Trueworthy v. Gelman Construction Co., 455 F.2d 1377, 1379 (D.C.Cir.1971). The evidence was not offered, or argued to the jury, as showing anything other than notice, and the jury was properly instructed that it could be considered for no other purpose. And, the jury’s rejection of exemplary damages also generally tends to indicate that it was not unduly inflamed with prejudice against Freightliner.
In the light of all these circumstances, we are unable to find reversible error in the matters complained of respecting the other lawsuits.
C. LASERE’S OPINION ON THE FREIGHTLINER STEP DESIGN.
As its final point of error, Freightliner argues that the district court abused its discretion in admitting the opinion testimony of plaintiffs’ witness, Robert Lasere, because'he was neither an expert nor properly qualified to express an opinion as a lay person under Rule 701.
Lasere was employed by Sigmor Truck Service as its service manager in charge of the daily maintenance of about sixty trucks, mostly Freightliners, and in charge of the major repairs and preventive maintenance on about 500 trucks altogether. He had roughly eighteen years of experience working on large trucks. Sigmor handled the maintenance of the truck involved in So-den’s accident, and Lasere was in charge of removing it from the scene.
Lasere testified, over Freightliner’s objection, to his observation of the damage to the fuel tanks on Soden’s truck and on two or three others he had seen following accidents. With respect to the other accidents (one rollover and one or two jackknifes of Freightliner trucks, none of which involved fires) he stated that when the wrecks were brought into the repair yard he had seen puncture holes in the fuel tanks at the location of the step brackets.- He testified that he had seen similar holes in the fuel tanks of Soden’s truck. He then gave his opinion that the step brackets were the [511]*511cause and that their design was dangerous. Finally, he described his decision, following Soden’s accident, to modify the step bracket, which secured the steps to the fuel tank, by sawing off its pointed ends. On appeal, Freightliner objects only to what it considers to be Lasere’s opinion testimony on the cause of the punctures and on the dangerousness of the step design.17 We hold that the district court acted within its discretion in admitting this evidence.
While the Sodens did not offer Lasere as an expert under Rule 703, he indisputably had very considerable practical experience and specialized knowledge. Rule 701, concerning lay opinions, provides:
“If the witness is not testifying as an expert, his testimony in the form of opinions or inferences is limited to those opinions or inferences which are (a) rationally based on the perception of the witness and (b) helpful to a clear understanding of his testimony or the determination of a fact in issue.”
This Court has noted three elements which must be present for lay opinion evidence to be admissible under Rule 701. Lubbock Feed Lots, Inc. v. Iowa Beef Processors, 630 F.2d 250 (5th Cir.1980). First, the witness must have “personal knowledge of the facts from which the ... opinion is said to derive.” Second, a “rational connection” must exist between the opinion and the facts upon which it is based; or put another way, “the opinion ... must be one that a normal person would form from those perceptions.” Finally, “the opinion ... must be helpful, either in understanding the testimony or in determining a fact in issue.” Id. at 263. The fact that an opinion goes to an “ultimate fact” does not necessarily preclude its admissibility under Rule 701. If these requirements are satisfied, a layman can under certain circumstances express an opinion even on matters appropriate for expert testimony. See Farner v. Paccar, Inc., 562 F.2d 518, 528-29 (8th Cir.1977); Randolph v. Collectramatic, Inc., 590 F.2d 844, 848 (10th Cir.1979). Cf. Scheib v. Williams-McWilliams Co., Inc., 628 F.2d 509, 511 (5th Cir.1980).18
Application of the Lubbock criteria to this case reveals that the admission of Lasere’s opinion testimony was not an abuse of discretion. First, contrary to Freightliner’s assertions, Lasere had suffi[512]*512cient personal knowledge of facts to form an opinion. Though he did not see any of the accidents firsthand, he soon after observed, in the normal course of performing his duties, holes or cuts in Freightliner fuel tanks near the location of the pointed step brackets, which was a sufficient basis for the opinion that the brackets were the cause of the holes. Second, Lasere’s conclusion was rationally supported and would have been apparent to a “normal person” in his position. His additional conclusion that this situation was “dangerous” was also rational. No great leap of logic or expertise was necessary for one in Lasere’s position to move from his observation of holes in Freightliner fuel tanks at the location of the step brackets, and presumably caused by them, to his opinion that the situation was dangerous. His testimony on this point did constitute an opinion which might have better been given by one more formally an expert; however, it had a strong basis both in his observation and in his experience. As Freightliner’s counsel admitted at oral argument, Lasere was a “practical expert” in the field of trucks, if not an expert in their design. Lasere’s testimony with respect to the dangerousness of the step brackets was also obvious, given the modification which he testified he made to them after all he had seen.
Finally, Lasere’s conclusion that the step brackets were causing punctures in Freight-liner fuel tanks was also helpful to the trier of fact in satisfaction of the third Lubbock criterion. It provided examples of one aspect of the design defect which the plaintiffs had alleged. During trial, the plaintiffs offered evidence that a diesel fire from the cab, not an oil fire from the trailer, had killed Soden, and that the weight of eighty pounds dropped on a step bracket from a height of one foot could puncture the fuel tank. Lasere’s testimony was a helpful link because it showed that an accident could also cause a puncture.
Thus, although Lasere’s opinion with respect to “dangerousness” may have been more properly made by one more formally an expert, given the particular facts of this case, we conclude that no reversible error occurred in its admission.
AFFIRMED.