FELDMAN, Justice.
This is a product liability case in which plaintiff appealed judgment entered for the defendant. The court of appeals affirmed.
Dart v. Wiebe Manufacturing, Inc.,
(No. 1 CA-CIV 5793, filed October 18, 1984). We accepted review in order to examine and settle .the law pertaining to strict liability claims involving alleged design defects.
See
Rule 23(f), Ariz.R.Civ.App.P. 17A A.R.S. We have jurisdiction under Ariz. Const, art. 6, § 5(3) and A.R.S. § 12-120.24.
FACTS
Wiebe Manufacturing, Inc. (defendant) manufactured a paper shredder and belt conveyor system used in the recycling of waste paper. The machine had no mechanical guards on the nip points at the rollers. Donald Dart (plaintiff) was an employee of Western Paper Stock Company, the owner of the machine. His employer instructed him to remove some paper wedged in a nip point in the conveyor. When he attempted to comply, his arm was pulled into the nip point where the flat belt passed over the roller and was torn off. He claimed the machine was defective and unreasonably dangerous because it was designed and distributed without safety guards which would have prevented the injury.
Plaintiff brought an action with separate counts of negligence and strict liability for the alleged design defect. The trial judge refused to give the separate instructions which plaintiff offered on each of these theories; insteád, he gave a single instruction to cover both of them. Plaintiff asserts on review that the instruction given effectively deprived him of his strict liability, design defect claim. He also argues that the separate instructions which he offered, but the trial judge refused, correctly embodied both the traditional approach to strict liability and the proper wording for negligence cases.
In affirming the judgment, the court of appeals correctly noted that the “adequacy and propriety of [the] instructions [given] revolve around the holding in
Brady v. Melody Homes Manufacturer,
121 Ariz. 253, 589 P.2d 896 (App.1978).” (Slip op. at 4.) The court construed
Brady
to hold that
There are two different bases of liability in design defect cases: ... if the product fails to perform as safely as an ordinary consumer would expect, ... then the manufacturer is liable under a strict liability standard. In addition, negligence is a possible theory of recovery if the issue is whether the manufacturer could have feasibly made the product safer.
Id.
Correctly noting that there was strong evidence of the feasibility of a safer design, the court of appeals reasoned, nevertheless, that plaintiff had “limited his own theories of recovery by couching his strict liability claim in the terminology of feasible [design] alternatives”
(Id.
at 6, 7). The basis for the court’s decision was that under
Brady
negligence principles are dispositive and strict liability concepts inapplicable when a plaintiff claims that a product design was improper and that the manufacturer should have adopted an alternative design. Thus, the ultimate conclusion was that the trial judge had ríbt erred in giving a jury instruction combining strict liability and negligence into a single theory.
While such a result may reflect the holding of
Brady v. Melody Homes, supra,
plaintiff argues that
Brady
and its proge
ny have misconstrued our previous opinions and have improperly merged strict liability and negligence theories in design defect cases. Accordingly, we turn first to an examination of the legal principles appropriate to design defect cases grounded on either negligence or strict liability theories.
DESIGN DEFECT TESTS
Although the doctrine of strict liability in tort imposes liability without proof of negligence, the law does not impose liability for every injury caused by a product. Liability exists only if the product was in a “defective condition unreasonably dangerous.” Restatement (Second) of Torts § 402A (1965) (hereinafter, Restatement §__) Thus, the central issue in a strict liability action is the definition to be given to the terms “defect” and “unreasonable danger.” Birnbaum,
Unmasking The Test For Design Defect: From Negligence [To Warranty] To Strict Liability To Negligence,
33 VAND.L.REV. 593, 597 (1980). Dean Wade, a leading commentator, calls this issue “the most vexing and pressing problem of products liability” law. Wade,
On Product “Design Defects” And Their Actionability,
33 VAND.L.REV. 551 (1980).
The Restatement definition resolves many cases; it defines a product in a “defective condition unreasonably dangerous” as one in a “condition not contemplated by the ultimate consumer, which will be unreasonably dangerous to him.” Restatement § 402(A) comment (g).
See also,
comment (i) (dangerous to extent beyond expectation of ordinary user or consumer). Ignoring the tautological problem,
this test provides a legal standard by which to measure the concept of “defect” and works well in manufacturing defect cases where, almost by definition, the product contains a danger which the manufacturer did not intend and the customer did not expect.
Brady v. Melody Homes Mfr.,
121 Ariz. at 256, 589 P.2d at 899. Indeed, the “consumer expectation” test arises from and expresses principles of implied warranty, the predecessor of strict liability in tort.
Wade, supra
at 552-53. Despite its virtues in the manufacturing defect setting, the consumer expectation test fails to provide an adequate legal standard in design defect cases.
Design defects present the most perplexing problems in the field of strict products liability because there is no readily ascertainable external measure of defectiveness. While manufacturing flaws can be evaluated against the intended design of the product, no such objective standard exists in the design defect context.
Caterpillar Tractor Co. v. Beck,
593 P.2d 871, 880 (Alaska 1979). Thus, while the consumer expectation test may sometimes work well in design defect cases
(Brady,
121 Ariz. at 256, 589 P.2d at 899), it provides no resolution for those cases in which “the consumer would not know what to expect, because he would have no idea how safe the product could be made.” Wade,
On the Nature of Strict Tort Liability for Products,
44 MISS.L.J., 825, 829 (1973);
Caterpillar Tractor v. Beck, 593
P.2d at 882 n. 35; Phillips,
The Standard For Determining Defectiveness In Products Liability,
46 U.CIN.L.REV. 101, 105 (1977). Further, in manufacturing defect cases consumer expectation can often be equated
with the manufacturer’s own expectations, but in design defect cases the problem of whether to measure consumer expectations by objective or subjective standards complicates the task of determining what is an “expectation.” Thus, concepts of absolute liability on the one hand or assumption of the risk and contributory negligence on the other may inadvertently be injected into the case. Birnbaum,
supra
at 599;
Caterpillar Tractor v. Beck, supra.
Some courts have devised a two-prong test for use in those design defect cases in which consumer expectation fails to provide an adequate standard. The first case to formulate this approach,
Barker v. Lull Engineering Co.,
20 Cal.3d 413, 143 Cal.Rptr. 225, 573 P.2d 443 (1978), holds that “defective design” exists either:
(1) if the product has failed to perform as safely as an ordinary consumer would expect when used in an intended or reasonable manner,
or
(2) if, in light of the relevant factors ..., the benefits of the challenged design do not outweigh the risk of danger inherent in such design.
20 Cal.3d at 418, 143 Cal.Rptr. at 228, 573 P.2d at 446 (emphasis supplied). The
Barker
test has been adopted in the case law.
See, e.g. Caterpillar Tractor v. Beck,
593 P.2d at 884;
Ontai v. Straub Clinic and Hospital,
66 Haw. 237, 659 P.2d 734 (1983).
Two years before
Barker
this court was faced with the problem of developing a satisfactory test for design defect cases. In
Byrns v. Riddell, Inc.,
113 Ariz. 264, 550 P.2d 1065 (1976), we stated that the concept of “unreasonable danger” was
... especially effective as a means of limiting the strict tort liability doctrine [where] the issue is the nature of the duty of a manufacturer with respect to safe design ... and in which there are serious questions as to the effect to be given harm-producing conduct or misuse on the part of the injured person.
113 Ariz. at 267, 550 P.2d at 1068 (quoting Annotation, 54 A.L.R.3d. 352, 358 (1973)). We expressly approved the risk/benefit analysis recommended by Dean Wade to determine if a design defect is “unreasonably dangerous.”
Id.
We believe that the
Barker
analysis is a logical refinement of our previous reasoning in
Byrns v. Riddell.
Where the consumer expectation test is inappropriate, the question of defective and unreasonably dangerous condition may be determined by applying Wade’s risk/benefit factors to decide whether “the benefits of the challenged design ... outweigh the risk of danger inherent in such design.”
Barker,
20 Cal.3d at 418, 143 Cal.Rptr. at 228, 573 P.2d at 446;
see also, Byrns,
113 Ariz. at 267, 550 P.2d at 1068.
Brady
strays from the path of strict liability at this point. Acknowledging that the
Barker
case “provides a springboard for the discussion of the product liability case based on design,” 121 Ariz. at 257, 589 P.2d at 900, the court concluded that the
Byrns-Barker
test and the Wade analysis adopted a negligence approach. 121 Ariz. at 258, 589 P.2d at 901. Further recognizing that
Byrns
had “quoted [with approval] a variant of the Wade definition” of unreasonable danger, the court of appeals stated in
Brady
that it was “not convinced that the Arizona Supreme Court has adopted the ramifications inherent in that definition.”
Id.
This conclusion is incorrect.
In adopting Dean Wade’s risk/benefit analysis for design defect cases, this court cited with approval and relied upon the case of
Dorsey v. Yoder Co.,
331 F.Supp. 753 (E.D.Pa.1971),
aff'd
474 F.2d 1339 (3rd Cir.1973). We stated:
The court in
Dorsey, supra,
subscribed to the following factor analysis prepared by Dean Wade to determine if a defect is unreasonably dangerous:
“(1) The usefulness and desirability of the product,
(2) the availability of other and safer products to meet the same need,
(3) the likelihood of injury and its probable seriousness,
(4) the obviousness of the danger,
(5) common knowledge and normal public expectation of the danger (particularly for established products),
(6) the avoidability of injury by care in use of the product (including the effect of instructions or warnings), and
(7) the ability to eliminate the danger without seriously impairing the usefulness of the product or making it undulyy expensive.”
Byrns v. Riddell,
113 Ariz. at 267, 550 P.2d at 1068 (quoting
Dorsey, supra,
at 760). This was not intended to adopt a negligence test for design defect cases. The purpose was exactly the opposite, and immediately after the above quotation we stated that
... The foregoing analysis is offered as an approach to the question of whether a defect is unreasonably dangerous.
Id.
We believe
Byrns
is good law today. The Restatement consumer expectation test is applied to the facts when the consumer may form an expectation:' In those cases in which consumer expectation fails to provide a complete answer,
Byrns
permits the use of a combined or alternative test which allows the jury to find that the product was “unreasonably dangerous” by applying the factors relevant to safe design. Use of these risk/benefit factors does not signal the abolition of strict liability in tort. It is, on the contrary, simply an alternative method of determining unreasonable danger.
We acknowledge that the risk/benefit analysis factors formulated by Dean Wade are also applicable to and should be used in negligent design cases. What then is the difference between a negligent design case and a strict liability design case? If there is none, then
Brady
may have been correctly decided and plaintiffs argument that the trial court erred in instructing on a single, “combined” theory is not well taken.
THE DISTINCTION BETWEEN NEGLIGENT DESIGN CASES AND STRICT LIABILITY DESIGN CASES
Brady
took the view that the risk/benefit analysis was essentially a negligence test because the court could “not conceive how” the fact finder could consider the risk/benefit factors “without considering whether the manufacturer’s conduct was reasonable in choosing the design utilized.”
Brady v. Melody Homes,
121 Ariz. at 259, 589 P.2d at 902. Thus, the court concluded “that where the harm-producing design does not fall within the Restatement [consumer expectation test], strict liability does not apply.”
Id.
The court reached its ultimate conclusion in the following words:
Rather, if liability is to be imposed on the grounds that the manufacturer could have feasibly made the product safer (we use the term “feasibly” here to denote a risk/benefit analysis), then the inquiry must be based upon whether the conduct of the manufacturer was reasonable, that is, standard principles of negligence.
Id.
We disagree with this conclusion.
We believe that there is a fundamental difference in the application of a risk/benefit analysis in a negligent design case and the same analysis in a strict liability design case. The difference is significant, for it shifts the central focus of the inquiry from the
conduct
of the manufacturer (negligence) to
the quality of the product
(strict liability). Negligence theory concerns itself with determining whether the conduct of the defendant was reasonable in view of the foreseeable risk of injury; strict liability is concerned with whether the product itself was unreasonably dangerous.
See, Salt River Project Agr. v. Westinghouse,
143 Ariz. 368, 376, 694 P.2d 198, 206 (1984);
Beach v. City of Phoenix,
136 Ariz. 501, 504, 667 P.2d 1316, 1319 (1983);
see also
Wade at 556, 566-69.
It is important to bear in mind that both the consumer expectation test and the risk/benefit analysis test “can embody strict liability because the manufacturer-designer may be held liable in situations where due or even utmost care would not have prevented the design hazard.” Henderson,
Strict Products Liability and Design Defects in Arizona,
26 ARIZ.L.REV. 261, 264 (1984). The true distinction, we believe, between negligent design cases applying the risk/benefit analysis and strict liability cases applying the same word formulation is the time frame in which this
determination is made. For a plaintiff to prove negligence he must prove that the designer or manufacturer acted unreasonably at the time of manufacture or design of the product. This test is “nothing more than the familiar negligence standard.”
Id.
at 265. Thus it is
... clear that the ‘risk-utility’ test is merely one of negligence when that test is confined to information or knowledge regarding the product
as of the date of manufacture-design as compared to a later date
such as the date of sale, accident or trial. To ... limit the jury’s consideration [in making the risk/benefit analysis] to that information known about the product, including reasonably foreseeable dangers, at the date of manufacture-design is to ask whether an ordinarily prudent manufacturer-designer would have produced and sold the product as it was produced and sold. Negligence by any other name is still negligence.
Id.
(emphasis supplied)
In a strict liability risk/benefit analysis, however, it is not the conduct of the manufacturer or designer which is primarily in question, but rather the quality of the end result; the product is the focus of the inquiry. The quality of the product may be measured not only by the information available to the manufacturer at the time of design, but also by the information available to the trier of fact at the time of trial. In
Byrns
we quoted with approval
“[t]he following test of “unreasonable danger”: ‘whether a reasonable manufacturer would continue to market his product in the same condition as he sold it to the plaintiff
with knowledge of the potential dangerous consequences the trial just revealed.
’ ”
113 Ariz. at 267, 550 P.2d at 1068 (quoting from
Dorsey v. Yoder,
331 F.Supp. at 759-60) (emphasis supplied). This “hindsight” test is generally recommended by the commentators, and by precedent, if the case is to be pursued in strict liability.
See, e.g.,
Keeton,
Product Liability and the Meaning of Defect,
5 ST. MARY’S L.J. 30, 38 (1973) ; Montgomery & Owen,
Reflections on the Theory and Administration of Strict Tort Liability for Defective Products,
27 S.C.L.REV. 803, 844 (1976);
Caterpillar Tractor Co. v. Beck, supra; Barker v. Lull, supra; Phillips v. Kimwood Machine Co.,
269 Or. 485, 525 P.2d 1033 (1974).
The hindsight test has been described as a “prudent manufacturer test” because the factfinder must evaluate the reasonableness of the manufacturer’s conduct. However, the test actually concerns itself with the quality of the product.
The problem with strict liability ... has been one of limitation. No one wants absolute liability____ To impose liability there has to be something about the article which makes it dangerously defective without regard for whether the manufacturer was or was not at fault____ A test for unreasonable danger is therefore vital. A dangerously defective article would be one which a reasonable person would not put into the stream of commerce
if he had knowledge of its harmful character.
The test, therefore, is whether the seller would be negligent if he sold the article
knowing the risk involved.
Strict liability imposes what amounts to constructive knowledge of the condition of the product.
Phillips v. Kimwood Machine,
269 Or. at 491, 525 P.2d at 1036 (emphasis in original footnotes omitted);
see also, Wilson v. Piper Aircraft Corp.,
282 Or. 61, 577 P.2d 1322 (1978);
Barker v. Lull, supra.
We believe, therefore, that the correct test was set forth in
Byrns v. Riddell, supra,
for determination of “unreasonable danger” in a design defect case in which the consumer expectation test cannot be applied or is uncertain. In a negligence case, the inquiry focuses on the reasonableness of the manufacturer’s choice of design
in light of the knowledge available at the time of manufacture. Under strict liability, however, knowledge of the “danger in fact” as revealed by the accident and the testimony at trial is imputed to the manufacturer. W. Prosser and W. Keeton, LAW OF TORTS, § 99 at 699 (5th ed. 1984). The question is whether, given the risk/benefit factors outlined in
Byrns
and any others which may be applicable, it was unreasonable for a manufacturer with such knowledge to have put the product on the market. If the question is answered in the affirmative, then the product is not only defective, but is unreasonably dangerous. If the question is answered in the negative, then the product is not unreasonably dangerous.
APPLICATION OF THE BRADY RULE
The court of appeals has generally avoided application of the
Brady
rule. For instance, in
Hohlenkamp v. Rheem Manufacturing Co.,
134 Ariz. 208, 655 P.2d 32 (App.1982), Division Two of the court acknowledged that if there was no consumer expectation and a risk/benefit analysis was applicable, then under
Brady
“the defect must be judged upon negligence principles using the standard of a reasonably prudent manufacturer.” 134 Ariz. at 211, 655 P.2d at 35. However, the court held that reasonable minds could differ with respect to whether there would have been consumer expectations with regard to the design of the product; therefore, the trial court had erred in refusing to submit the issue of strict liability to the jury. 134 Ariz. at 212, 655 P.2d at 36. The same court refused to apply
Brady
in
Moorer v. Clayton Manufacturing Corp.
128 Ariz. 565, 627 P.2d 716 (App.1981), a case which also involved a claim of improper design for failure to provide a safety guard at the nip-point of a machine. Holding that it “need not state here whether we agree with
[Brady’s]
apparent retreat from the doctrine of strict liability,” (128 Ariz. at 568, 627 P.2d at 719), the court stated that a jury could find that the lack of a safety guard fell below the reasonable expectations of an ordinary consumer and concluded that plaintiff had established a strict liability claim. In
Turner v. Machine Ice Co.,
138 Ariz. 329, 674 P.2d 883 (App.1983), a panel of Division One followed
Byrns,
properly concluding that although the language in a risk/benefit approach “employs negligence terminology” it is actually a method of testing for unreasonable danger and avoiding imposition of absolute liability. 138 Ariz. at 333, 674 P.2d at 887. The court held, therefore, that the trial judge had erred in directing a verdict against plaintiff on a strict liability claim based upon improper design for omitting a sheer-point safety guard on a machine. 138 Ariz. at 335, 674 P.2d at 889.
The foregoing discussion indicates that the court of appeals has struggled with
Brady.
We therefore explicitly hold that
Brady
is disapproved insofar as it indicates that a design defect case in which the risk/benefit analysis is employed must be decided on negligence principles. Such cases should, if possible, be decided upon the consumer expectation test.
Hohlenkamp v. Rheem Manufacturing, supra.
If that test is inapplicable because the ordinary consumer would not form an expectation with regard to the relevant safety feature, then the case must proceed under the risk/benefit factor analysis discussed in
Byrns v. Riddell, supra.
In close cases it may be necessary for the court to give both instructions. Employment of this methodology does not transform the case into one of negligence. The case may proceed on the theory of strict liability, even though the jury, by using the risk/benefit factors, focuses upon the reasonableness of the manufacturer’s conduct in putting the product on the market. However, in so doing it is instructed to impute to the manufacturer knowledge of the danger that has been revealed subsequent to distribution of the product.
The question is
whether a reasonable manufacturer, with knowledge of such dangers, nevertheless would have put the product on the market. With these principles in mind, we turn now to the case before us.
THE INSTRUCTIONS
This case illustrates the confusion created by
Brady.
For reasons not disclosed in the record, the trial court assumed that the failure to provide a safety guard for the nip point was a matter beyond ordinary user or consumer expectations. Plaintiff has not assigned this as error.
Cf. Moorer v. Clayton Manufacturing Co.,
128 Ariz. at 568, 627 P.2d at 719.
In any event, the trial court sent the case to the jury with a risk/benefit analysis instruction. It attempted to tailor the instruction to accommodate the
Brady
view that by so doing the case was changed into a negligeñce claim. The relevant portions of the instruction follow:
Plaintiff claims that defendant’s product was defective and unreasonably dangerous in that it was negligently designed. Negligent design can be described as the manufacturer failing to do that which an ordinarily prudent manufacturer would do or not doing something which an ordinarily prudent manufacturer would do____
Negligent conduct ... is that conduct which falls below the standard of care required of a manufacturer of conveyors. ... A manufacturer is negligent if he fails to act as an ordinarily careful manufacturer would act ... thus, if better testing or design was available and would have eliminated an unreasonable danger the manufacturer is expected to have known and used that method as a matter of due care. Negligence, however, is not established with a simple showing that a better design could be built [but only] where a showing is made that the manufacturer was unreasonable in not adopting the better design.
In determining whether or not the product was unreasonably dangerous, you may consider the following factors____
The instruction ended with a recitation of eleven factors for the jury to consider in determining the issues. Most of the factors were taken from the list provided in
Byrns v. Riddell, supra;
the rest presumably came from
Brady.
It is obvious that the instruction given combined negligence and strict liability terminology. We do not, of course, look solely to the terminology used but also to the substance of the instruction. We conclude that in substance the instruction is a negligence instruction because it requires the jury to decide whether the manufacturer has “[failed] to act as an ordinary careful manufacturer would act under the circumstances.” Earlier, the instruction had informed the jury that negligent design was to be determined by measuring defendant’s conduct against “that [of] an ordinarily prudent manufacturer.” The focus of this instruction, therefore, is not upon the quality of the product but the quality of the defendant’s conduct.
See Rossell v. Volkswagen,
147 Ariz. 160, 709 P.2d 517, 524 n. 3 (1985). Again, this is the essence of negligence rather than strict liability.
Barker v. Lull Engineering,
20 Cal.3d at 418, 143 Cal.Rptr. at 229, 573 P.2d at 447.
The court of appeals held that the instruction was not error. (Slip op. at 5-8.) The court acknowledged that plaintiff had “arguably presented evidence in support of both” the consumer expectation test and risk/benefit analysis
(Id.
at 5), but concluded that the “primary thrust at trial was to show that design alternatives were feasible.” Therefore, the court concluded, plaintiff “injected a negligence theory into the strict liability instruction he submitted.” Thus, “in effect, Dart limited his own theories of recovery by couching his strict liability claim in the terminology of feasible alternatives.” (Slip op. at 6.) We do not agree with this analysis. It is a product of
Brady
and depends for its validity on
Brady’s
erroneous conclusion that use of the risk/benefit analysis transforms the case into a negligence claim.
We believe plaintiff’s proposed instructions embodied a correct statement of the law. Plaintiff first proposed an instruction that the same set of facts may support “more than one theory of recovery” and that plaintiff had two theories, one based upon strict liability and the other on negligence. Plaintiff then submitted the following (Requested Instructions Nos. 2, 3 and 4) to apply to the strict liability claim:
(2) A party who manufactures or sells a product in a defective condition unreasonably dangerous to its user is liable if....
******
(3) In determining whether a defect in a product is unreasonably dangerous, you should consider ... [listing the seven factors enumerated in
Byrns v. Riddell, supra].
Plaintiff’s last instruction on strict liability theory was based upon
Dorsey v. Yoder, supra
and read as follows:
(4) In the instant case, you may conclude that the [product] was unreasonably dangerous if you determine that a reasonable manufacturer with knowledge of the potential dangerous consequences the trial has just revealed would not continue to market its product in the same condition in which it was sold.
Plaintiff then requested instructions on his negligent design claim. These essentially consisted of an instruction informing the jury that “a person is negligent if he fails to act as an ordinarily careful person would act under the circumstances,” followed by an instruction telling the jury that “in determining whether the defendant was negligent, you should consider” the same seven factors previously given the jury and extracted from
Byrns v. Riddell, supra.
We believe these instructions are basically correct. Requested Instructions Nos. 2, 3 and 4 were strict liability instructions. Plaintiff did not foreclose his use of a strict liability theory by his failure to request an instruction on consumer expectation, though under
Moorer, supra
he may well have been entitled to such an instruction.
We are hesitant to reverse any case on an instruction issue.
See, e.g., Petefish v. Dawe,
137 Ariz. 570, 577, 672 P.2d 914, 921 (1983). However, where the challenged instructions cut to the very heart of the case and misapply the applicable legal theories, the error must be considered prejudicial.
Chavez v. Pima County,
107 Ariz. 358, 488 P.2d 978 (1971). We hold that the trial court erred in refusing the instructions which the plaintiff requested; by giving the hybrid negligence instruction the trial court committed prejudicial error because the legal theories properly advanced by plaintiff were misapplied and miseharacterized. Plaintiff made timely objection to the court's instruction and to the court’s refusal of his requested instructions. Accordingly, we must reverse. Our disposition of the case makes it unnecessary to consider other issues raised by the plaintiff or the opinion of the court of appeals.
The opinion of the court of appeals is vacated. The judgment of the trial court is reversed. The case is remanded for a new trial and for further proceedings not inconsistent with this opinion.
GORDON, V.J.C., and HAYS and CAMERON, JJ., concur.
HOLOHAN, C.J., dissents.