AMENDED OPINION
FERGUSON, Circuit Judge:
Plaintiffs, the Oregon Natural Resources Council, Oregon Guides and Packers Association, Rogue Flyfishers, Inc., and Rogue River Guides Association (“Resources Council”), appeal the denial of an injunction to stop construction by the United States Army Corps of Engineers (“Corps”) of a dam on Elk Creek in Southern Oregon. The Resources Council contends that the Corps’ final supplemental environmental impact statement (“EIS”) for the dam project fails to comply with the National [1492]*1492Environmental Policy Act of 1969, Pub.L. No. 91-190, 83 Stat. 853 (1970) (codified at 42 U.S.C. § 4332) (“NEPA”). We affirm in part and reverse in part.
I.
Elk Creek is a tributary of Oregon’s Rogue River. The Rogue was one of eight rivers in the nation designated as “wild and scenic” in the Wild and Scenic Rivers Act of 1968, Pub.L. No. 90-542, 82 Stat. 906 (1968) (codified at 16 U.S.C. § 1271 et seq.). The Rogue is known nationally for its white-water rafting and fly-fishing.
In 1962, Congress authorized construction of three dams to control flooding of the Rogue River Basin: Elk Creek Dam, Applegate River Dam, and Lost Creek Dam. The latter two dams are completed. The Elk Creek project consists of a proposed concrete dam 249 feet high and 2,580 feet long. The reservoir behind the proposed dam would collect runoff from most of Elk Creek’s watershed and cover 1,290 acres of land.
In 1971, the Corps filed a final EIS for the Elk Creek Project. In 1975, after studying the projected effects of the dam on turbidity1 in the Rogue River, the Corps filed a draft supplemental EIS. Construction of the project was suspended until completion of further water quality studies.
In 1980, the Corps filed a revised draft supplemental EIS, which included turbidity and temperature projections based on the Corps’ observations of turbidity caused by the Lost Creek Dam during the year 1977 and the Corps’ computer simulation models. Several state and federal agencies criticized the Corps’ turbidity analysis on the ground that rain runoff levels during 1977 were the lowest on record in the area. The Corps placed these critical comments, and its responses to them, in a separate section, and released the aggregate as its final supplemental EIS.
The Resources Council2 then brought this action in federal district court for a temporary restraining order, a preliminary injunction, and a permanent injunction to stop construction of Elk Creek Dam. The Resources Council argued that the final supplemental EIS failed in several ways to comport with the requirements of the NEPA and regulatory guidelines of the Council on Environmental Quality. The Resources Council also contended that the Corps was required to prepare a new supplemental EIS because of new information. The district court found the final supplemental EIS adequate and denied the requested relief. The Corps has since awarded contracts for relocation of roads and utilities and for construction of the dam.
II.
The Resources Council contends that the final supplemental EIS violates the NEPA procedural standards and the Council on Environmental Quality guidelines in six respects. The district court held that the environmental impact statement comported with these standards. We review de novo the district court’s determination of whether the EIS is reasonably thorough in discussing the significant aspects of the probable environmental impact of the proposed agency action. Enos v. Marsh, 769 F.2d 1363, 1372 (9th Cir.1985) (applying a “rule of reason”). Purely factual findings are, of course, reviewed for clear error. Id.
We thus endeavor to determine whether the final supplemental EIS satisfies the two purposes of an EIS: (1) to provide decisionmakers with enough information to aid in the substantive decision whether to proceed with the project in light of its environmental consequence; and (2) [1493]*1493to provide the public with information and an opportunity to participate in gathering information. Citizens for a Better Henderson v. Hodel, 768 F.2d 1051, 1056 (9th Cir.1985); California v. Block, 690 F.2d 753, 761 (9th Cir.1982) (the “form, content and preparation [of the EIS] foster both informed decision-making and informed public participation”); 40 C.F.R. § 1502.1 (purpose of EIS is to “provide full and fair discussion of significant environmental impacts and ... [to] inform the decisionmakers and the public of the reasonable alternatives which would avoid or minimize adverse impacts ... ”).
First, the Resources Council contends that the final supplemental EIS is deficient because it fails to discuss adequately ways to mitigate adverse environmental impacts of the project. Specifically, the Resources Council claims that the mitigation plan for wildlife contains neither a detailed analysis of mitigation measures nor an explanation of how effective those measures would be. We agree, and reverse for further proceedings on this issue.
An EIS must include a discussion of measures to mitigate adverse environmental impacts of the proposed action. 40 C.F.R. § 1502.16(h). As long as significant measures are undertaken to mitigate the project’s effects, the measures need not compensate completely for adverse environmental impacts. Friends of Endangered Species, Inc. v. Jantzen, 760 F.2d 976, 987 (9th Cir.1985). The mere listing of mitigation measures, however, is insufficient to satisfy the NEPA requirements. Northwest Indian Cemetery Protective Ass’n v. Peterson, 795 F.2d 688, 697 (9th Cir.1986). Moreover, the EIS must analyze the mitigation measures in detail and explain the effectiveness of the measures. See id.
Here, the mitigation plan for wildlife is not yet fully developed. The mitigation plan, published in 1980, states: “Measures to compensate project-caused loss of wildlife habitat associated with reservoir construction will be developed, based upon the results of the wildlife compensation plan currently underway at Applegate Project and recommendations of Oregon Department of Fish and Wildlife and the U.S. Fish and Wildlife Service.” No mitigation plan had been finalized as of January 1986. We fail to see how mitigation measures can be properly analyzed and their effectiveness explained when they have yet to be developed.
Moreover, the mitigation measures set forth in the final supplemental EIS are inadequate to satisfy the Council on Environmental Quality guidelines. The mitigation plan for wildlife, in its entirety, states:
The compensation [for project-caused loss of wildlife habitat] would be accomplished by managing selected lands to increase the quality of habitat and therefore its carrying capacity for wildlife. The management would consist of a number of habitat manipulative techniques in conjunction with other land uses, landscaping at the project, and development of recreation sites. Specific habitat development measures would be oriented toward development of palatable browse plants, creation of “edge,” maximizing foliage height diversity in certain areas, and creation of snags in the reservoir. This compensation plan will be developed in coordination with Federal and state resource agencies.
This plan lists general measures to mitigate the impact of the project on wildlife. The plan refers to “habitat manipulative techniques,” but fails to specify which techniques will be used. It also fails to identify any of the “specific habitat development measures” that would be utilized to develop palatable browse plants, create “edge,” maximize foliage height diversity, or create snags in the reservoir. More important, there is no analysis of the mitigation measures listed, or any estimation of how effective the measures will be.
The importance of the mitigation plan cannot be overestimated. It is a determinative factor in evaluating the adequacy of an environmental impact statement. Without a complete mitigation plan, the decisionmaker is unable to make an informed judgment as to the environmental [1494]*1494impact of the project—one of the main purposes of an environmental impact statement. See Citizens for a Better Henderson, 768 F.2d at 1056.
Because the wildlife mitigation plan here merely lists measures to be used and includes neither an analysis nor an explanation of effectiveness, it is inadequate to satisfy the NEPA or Counsel on Environmental Quality mitigation guidelines. See Northwest Indian Cemetery Protective Ass’n, 795 F.2d at 697.
III.
The Resources Council also contends that the Corps violated the NEPA by failing to prepare a new supplemental EIS taking into account new information acquired since the 1980 EIS. We agree, and remand for further proceedings on this issue as well.
A federal agency has a continuing duty to gather and evaluate new information relevant to the environmental impact of its actions after release of an EIS. Stop H-3 Ass’n v. Dole, 740 F.2d 1442, 1463 (9th Cir.1984), cert. denied, 471 U.S. 1108, 105 S.Ct. 2344, 85 L.Ed.2d 859 (1985). The agency must supplement the EIS if “[t]here are significant new circumstances or information relevant to environmental concerns and bearing on the proposed action or its impacts.” 40 C.F.R. § 1502.9(c)(1)(h); see Enos, 769 F.2d at 1373. When new information comes to light, the agency must consider it, evaluate it, and make a reasoned determination whether it is of such significance as to require implementation of formal NEPA filing requirements. Stop H-3 Ass’n, 740 F.2d at 1463-64.
We will uphold an agency’s decision not to supplement an environmental impact statement in light of new information if the decision is reasonable. Enos, 769 F.2d at 1373.
Reasonableness depends on the environmental significance of the new information, the probable accuracy of the information, the degree of care with which the agency considered the information and evaluated its impact, and the degree to which the agency supported its decision not to supplement with a statement of explanation or additional data.
Stop H-3 Ass’n, 740 F.2d at 1464 (citation omitted). Applying the Stop H-3 Ass’n standard,3 we conclude that subsequent to the preparation of the final supplemental EIS, significant new information came to light. We further conclude, based upon the proceedings in the district court, that this information is probably accurate. The Corps failed to exercise the proper degree of care in evaluating the impact of the new information and failed to properly support its decision not to supplement the EIS with an explanation or additional data. Thus, a new supplemental EIS should have been prepared.
Two studies provided significant new information regarding the environmental impact of the project. The Oregon Department of Fish and Wildlife concluded that the project could result in decreased survivability of chinook salmon, higher turbidity, increased disease potential in fish, and decreased prospects for lure- and fly-fishing.4 The United States Soil Conservation Service found a greater turbidity potential than indicated by the final supplemental EIS. These studies contained the significant new information contemplated by Stop H-3 Ass’n.
The final draft of the Fish and Wildlife Department study, dated August 5, 1985, was based on the impact of Lost Creek Dam on salmon and steelhead populations [1495]*1495downstream. Department scientists observed decreased survival of chinook salmon, and speculated that it was a result of increased water temperatures caused by the release of warmer water from the Lost Creek Reservoir. Based on Lost Creek Dam’s impact on survivability, the Department believed that conditions from Elk Creek Dam also would affect negatively chinook salmon survivability. Department scientists concluded that, even though their study was not conclusive, this phenomenon should be studied further before Elk Creek Dam is built.
With regard to higher turbidity, information was based on a study of logging and road-building activity in the Elk Creek watershed since 1980. The logging and road-building has caused soil disturbance. Fish and Wildlife Department scientists believe that this activity will result in higher turbidity than that reported in the final supplemental EIS. Consequently, they suggest additional studies on turbidity potential before the dam is built.
As to disease potential in fish, Fish and Wildlife Department statistics indicate that in 1979, following the construction of Lost Creek Dam, there was a 76% mortality rate in fall chinook salmon populations due to disease and a 32% mortality rate in 1980. Although there has been almost no mortality since 1981, Department scientists concluded there may be a connection between dam closure and the potential for increased disease among fish.
Finally, the Fish and Wildlife Department was concerned about higher flows in the Rogue River that could decrease the effectiveness of lure- and fly-fishing along the river, including the wild and scenic portion. The Department wanted to conduct studies to evaluate this concern, but the Corps reduced the Department’s budget, making further study impossible. The Corps evaluated the Department study and concluded that the new information was not significant enough to warrant a new supplemental EIS.
The Soil Conservation Service soil study also revealed significant new turbidity information tending to corroborate the Department study. The Soil Conservation Service study indicates that the Elk Creek watershed has substantially greater turbidity potential than previously reported in the EIS. Uncontradicted testimony showed that the soil content in the watershed is different than indicated in the final supplemental EIS. The Corps conducted no studies after receiving this new information because it had investigated erosion potential in 1979 and found nothing to substantiate the Department’s concerns over the soil study. Thus, the Corps concluded that no significant new information had been discovered.
The district court found that the Corps’ decision not to release a new supplemental EIS was reasonable. Like the district court, we review the agency’s decision for reasonableness. See Enos, 769 F.2d at 1373. We conclude otherwise. The Department presented new information as a result of its study of Lost Creek Dam’s impact on the river. Although the evidence is not conclusive, it presented a legitimate concern about decreased survivability of fish and the potential for higher turbidity based on information from the two studies that was not available at the time the final supplemental EIS was published. The evidence clearly is environmentally significant under the Stop H-3 Ass’n test.
The information must be more than environmentally significant. It must be “probably accurate” as well. Stop H-3 Ass’n, 740 F.2d at 1464. Based on our review of the experts’ testimony and data, we believe that, at the very least, some of the proffered information is probably accurate. Although the Corps had two independent experts evaluate the Fish and Wildlife Department’s study, and these experts were critical of many of the conclusions reached, the experts also agreed with significant portions of the study. The Corps did not dispute that there was increased logging and road-building activity in the Elk Creek watershed since 1980. The Fish and Wildlife Department is concerned that this activity will result in increased turbidity.
The Corps also did not respond to the study’s conclusions concerning increased [1496]*1496fish mortality from epizootic diseases.5 The information admittedly was not conclusive. Yet, the study raised legitimate concerns regarding the relationship between dam construction and epizootic diseases. The Corps should have considered the information, evaluated its impact and, at the very minimum, supported its decision not to supplement with a statement of explanation or additional data. Id. The Corps does not appear to have exercised the proper degree of care regarding the new information. The Corps also has not supported its decision with a statement of explanation or additional data.
The testimony below also reveals that the information provided by the Soil Conservation Service was never considered by the Corps, even when the Corps issued its Supplemental Information Report in September 1985. In response to questions submitted by the Council of Environmental Quality (CEQ) in 1981, the Corps stated that soils in the Elk Creek Watershed erode more easily than those in the Lost Creek Watershed and therefore produce a higher volume of suspended material per unit of stream runoff and require a longer time to settle out. The Corps concluded that this factor was considered in the modeling effort. Later, in response to commentary (identified as “Letter 14,” attached to the final supplemental EIS) submitted in response to the draft EIS, the Corps stated that morphological factors cannot be incorporated into the WESTEX model6 but are qualitatively included in the overall analysis of the potential turbidity of the watershed.
However, when presented with evidence showing that the soil content in the Elk Creek watershed was different from that indicated in the final supplemental EIS, the Corps chose not to respond. Table Four of the final supplemental EIS indicates that erosion potentials differ with soil associations. Further, the EIS states that turbidity is “generally caused by soil particles in the water column which are washed downstream in the runoff from the watershed.” Soil type and turbidity may not be directly related, as the Corps contends. Yet, the data reveals that soil type is related to erosion potential which, in turn, affects turbidity. The Corps took this into account earlier in its analysis and should do so again using the new information. Even if this data cannot be included in the WES-TEX model, the Corps should include it qualitatively in its “overall analysis” of turbidity.
These are significant matters, the accuracy of which is not disputed. The Corps did not address adequately these issues. Thus, applying the final two factors of Stop H-3 Ass’n, we conclude that the Corps must prepare a supplemental EIS to address these issues.7
IV.
The Resources Council also contends that the Corps was required by the CEQ “worst case” regulation to research further the effects of Elk Creek Dam on turbidity in the Rogue River during high, average, and normal-low rainwater runoff years.
The worst case analysis regulation, 40 C.F.R. § 1502.22, requires that when there are gaps in relevant information or scientific uncertainty with regard to significant adverse environmental effects, the agency must make it clear that the information is lacking or that uncertainty exists. If information is lacking that is essential to a reasoned choice and the cost of obtaining it is not exorbitant, the agency must obtain [1497]*1497the information and include it in the statement or include a worst case analysis.8
The Oregon Department of Fish and Wildlife formally criticized the Corps’ conclusion that Elk Creek Dam would have little effect on temperature and turbidity on the Rogue. The state agency specifically disagreed with the Corps’ finding that Elk Creek would make only a small contribution to total flow in the Rogue River and commented that there may be periods when Elk Creek provides a significant percentage of the total flow of the Rogue. The Corps, in responding to the state agency’s comment (identified as “Letter 6,” attached to the final supplemental EIS), acknowledged that Elk Creek could contribute up to 25% of the flow in the Rogue during high flow periods. Thus, the Corps’ assessment of turbidity due to Elk Creek Dam is subject to uncertainty. This uncertainty is disclosed in the final supplemental EIS.
The district court concluded that the Corps complied with the worst case analysis requirement because the final supplemental EIS included comments that exposed uncertainty, if uncertainty did in fact exist. Yet, exposing uncertainty is not enough. Save Our Ecosystems v. Clark, 747 F.2d 1240, 1244 n. 5 (9th Cir.1984), requires that the Corps include a worst case analysis in the EIS when there are gaps in the information engendering scientific uncertainty which the agency determines to be important and when the necessary information cannot be obtained because of scientific impossibility or because the costs of obtaining it would be exorbitant. See also Oregon Environmental Council v. Kunzman, 817 F.2d 484, 495 (9th Cir.1987). The Corps did not include a worst case analysis and did not conduct any further research with regard to the water flow contribution from Elk Creek. The Corps should pursue one of these alternatives.
Additionally, since the Corps is now required to consider new information brought forth by the two studies, the Corps should take this information into account in its decision concerning a worst case analysis. If uncertainty remains at that time, the Corps should prepare a worst case analysis or conduct further research with regard to the new information.
V.
The Resources Council further contends that the Corps unreasonably limited the scope of the final supplemental EIS by failing to consider the cumulative effects of all three dam projects, Lost Creek Dam, Applegate Dam, and Elk Creek Dam, the first two of which already have been completed. The Resources Council contends that the Corps should address the cumulative effects of the dams in a single EIS because the three components of the Rogue River Basin Project are connected.9
The “cumulative impact” regulation requires the Corps to evaluate “the incremental impact of the action when added to other past, present, and reasonably foreseeable actions.” 40 C.F.R. § 1508.7. Although the CEQ guidelines require that “cumulative actions” be considered together in a single EIS, 40 C.F.R. § 1508.25(a)(2), and “cumulative actions” consist only of “proposed actions,” this does not negate the requirement of 40 C.F.R. § 1508.7 that the Corps consider cumulative impacts of [1498]*1498the proposed actions which supplement or aggravate the impacts of past, present, and reasonably foreseeable actions.
The district court correctly ruled that a separate EIS was not required for the entire project. Only proposed actions must be discussed in a separate EIS. 40 C.F.R. § 1508.25(a)(2). The court also concluded that the final supplemental EIS adequately described and analyzed the impacts in terms of the effects of the entire project. The court was convinced that the Corps took a hard look at the cumulative impacts in the EIS.
We disagree with the district court that the Corps took a hard look at the cumulative environmental impacts. Plaintiffs point out two examples of the Corps’ failure to consider the cumulative impacts. First, although the final supplemental EIS conceded that the overall basin project would increase the turbidity of the Rogue River, it concluded that there would be no major adverse effect on fish production from increased turbidity. In reaching this conclusion, however, it considered only the turbidity created by the individual project. Similarly, in response to a comment made by the United States Environmental Protection Agency suggesting that the EIS should discuss the effects of Lost Creek Dam on downstream water quality, the Corps responded that the Elk Creek EIS was not the proper place to discuss the effects of Lost Creek Lake. Insofar as the effects of the Lost Creek Dam were necessary to a complete presentation of the cumulative impact of construction of Elk Creek Dam, we disagree.
The synergistic impact of the project should be taken into account at some stage, and certainly before the last dam is completed. In Kleppe v. Sierra Club, 427 U.S. 390, 96 S.Ct. 2718, 49 L.Ed.2d 576 (1976), the Court specifically noted that “an agency could approve one pending project that is fully covered by an impact statement, then take into consideration the environmental effects of that existing project when preparing the comprehensive statement on the cumulative impact of the remaining proposals.” Id. at 415 n. 26, 96 S.Ct. at 2732 n. 26.
While a separate EIS is not required, the Corps should consider the impact of Elk Creek Dam in conjunction with Lost Creek Dam and, if appropriate, Applegate River Dam. The Corps is building Elk Creek Dam in an area that already has two dams. It must consider the area as it finds it and take into consideration the cumulative impact of the basin project.
VI.
We affirm the district court’s rulings on the Resources Council’s remaining three challenges. First, the Resources Council contends that the final supplemental EIS gave no meaningful response to comments critical of the Corps’ draft supplemental EIS, because these comments were not integrated into the body of the final supplemental EIS. The Resources Council relies on a CEQ regulation that requires discussion “at appropriate points in the final [EIS of] any responsible opposing view.” 40 C.F.R. § 1502.9(b). The Resources Council argues that the only “appropriate points” in an EIS for discussion of opposing views are in the body.
We first note that section 1502.9(b) does not define “appropriate points.”10 We must therefore examine that guideline and the Corps’ final supplemental EIS in light of the purposes of an environmental impact statement of providing complete information to the decisionmaker and the public.
We find that an agency may place responsible opposing views in a separate “comments and responses” section when, as here, many of the critical comments prompted revisions in the body, the Corps discussed in the body all of the environmental problems to which the comments [1499]*1499were addressed, and the Corps provided thoughtful and well-reasoned responses to most of the critical comments. We therefore hold that the Corps sufficiently complied with the purposes of the NEPA in its treatment of opposing views, even if the most “appropriate points” for addressing those views would be in the body.11
Second, the Resources Council contends that the final supplemental EIS is inadequate because it fails to discuss, as required by 40 C.F.R. § 1502.16(c), possible conflicts between the proposed action and other federal policy objectives for the affected area. Specifically, the Resources Council argues that the agency failed to analyze the effect of the dam on the wild and scenic portion of the Rogue River, located fifty-seven miles downstream. An EIS need contain only a reasonably thorough discussion of the significant aspects of the probable environmental consequences of a proposed action. Trout Unlimited v. Morton, 509 F.2d 1276, 1283 (9th Cir.1974)).
In this case, the United States Forest Service commented that the final supplemental EIS should address the effect of the project on the wild and scenic portion of the river. The Corps responded that, based on its conclusion that the impact on the immediate downstream portion of the river would be insignificant, the impact on the wild and scenic portion fifty-seven miles downstream also would be insignificant. The agency assumed that the reservoirs’ impacts diminish with distance.
Applying this court’s “rule of reason,” see Enos, 769 F.2d at 1372, the agency logically concluded that the impact from the project on the wild and scenic portion of the river would be insignificant.12 Assuming the agency reaches the same con-elusion concerning the immediate downstream portion of the river after it takes into consideration the new information, its decision concerning the wild and scenic portion of the river will stand.
Third, the Resources Council contends that the final supplemental EIS is defective because the Corps placed a cost-benefit analysis of the Elk Creek Project using a 3V4% discount rate in the body, but relegated an analysis based on a more realistic llk% rate to an appendix. The Resources Council argues that the placement of these analyses misled the decisionmaker as to the economic value of the project.
An EIS is not required to contain a formal, mathematically expressed cost-benefit analysis. California v. Block, 690 F.2d at 764. Nevertheless, if such an analysis is included, it should not be misleading. See Johnston v. Davis, 698 F.2d 1088, 1094 (10th Cir.1983).
We find that the Corps’ use of a 3V4% discount rate or its placement of the 7V2% cost-benefit analysis in an appendix was not misleading. The Water Resources Development Act of 1974, Pub. L. No. 93-251, § 80, 88 Stat. 34 (1974) (codified at 42 U.S.C. § 1962d-17(b)), required use of the 3V4% discount rate. Johnston, 698 F.2d at 1092-95. The Corps’ inclusion of the analysis based on the more realistic 7¥2% comported with CEQ guidelines. See 40 C.F.R. § 1502.23 (cost-benefit analysis “shall be incorporated by reference or appended to the statement”). Moreover, although the purposes of an EIS might have been better served if the 3V4% analysis in the body had made reference to the 7V2% analysis in the appendix, the inclusion of the latter analysis sufficiently cured whatever misleading [1500]*1500effects resulted from use of the 3% percent rate. Cf. Enos, 769 F.2d at 1375 n. 14.
VII.
We conclude that the Corps’ mitigation report is inadequate, and that the Corps should file a new supplemental EIS because of new information obtained. The Corps also must prepare a worst case analysis or conduct further research with regard to the waterflow contribution from Elk Creek. The Corps also should conduct a worst case analysis based on the new information submitted in the two studies, if necessary. Finally, although the Corps need not prepare a separate statement addressing the cumulative impact of the dams, it should supplement the statement and address the cumulative impact of Lost Creek and Elk Creek Dams and, if appropriate, Applegate River Dam. We therefore remand to the district court for entry of appropriate injunctive relief in accordance with this opinion.
AFFIRMED IN PART, REVERSED IN PART, AND REMANDED.