with whom Justice Blackmun and Justice Stevens join, concurring in part and dissenting in part.
I join Part III of the Court’s opinion, granting the petitions to intervene in this action filed by the Fort Mojave, Colorado River, Chemehuevi, Cocopah, and Quechan Tribes (collectively, the Tribes). I also agree with the basic premise of Part IV of the Court’s opinion that in Article IX of our 1964 decree, 376 U. S. 340, 353, we retained the power to reconsider our quantification of the Tribes’ reserved water rights, as set out in Article 11(D) of the 1964 decree, id., at 343-345. See ante, at 618. I part company with the Court, however, in its refusal to exercise that power, given the unique circumstances of this litigation and the timing of the Tribes’ and United States’ motions. In addition, I find inexplicable the Court’s decision to sustain the exceptions of Arizona, California, and the California agencies (hereinafter States) to the Special Master’s proposed solution to the boundary lands controversy.
I
The so-called “omitted” lands are irrigable areas, within the Tribes’ reservations, which the United States failed to identify during the extensive proceedings before Special Master Rifkind that preceded our 1964 decree. The fact that irrigable lands were not called to the attention of the Master or the Court is significant because the Master and the Court held that the amount of water which the Tribes were entitled to divert from the mainstream of the Colorado depended on the number of “irrigable acres” within each reservation. 373 U. S. 546, 601 (1963); Report of Special Master Rifkind 263-265 (hereinafter Rifkind Report). Although the States vociferously dispute exactly how much of the omitted lands *643are in fact irrigable, they do not dispute two facts critical to the question now before the Court. First, even under the States’ legal theories a substantial portion of the omitted lands are irrigable — at least 18,500 acres, see Report of Special Master Tuttle 109, 125 (hereinafter Tuttle Report) — and would have supported an award of additional diversion rights in our 1964 decree had they been identified at that time. Second, the United States completely failed to present evidence regarding the irrigability of these lands until after the Tribes sought leave to intervene in these proceedings in 1977.
There are strong arguments for correcting the quantifica-tions of the Tribes’ diversion rights in the 1964 decree, to include the amounts of water that could be used economically to irrigate the omitted lands. As this litigation now stands, the considerations of finality are not so strong, nor the interests of justice so weak, as the Court would have them. The system contemplated by our 1964 decree for allocating the waters of the Colorado River’s Lower Basin has yet to become final, either as a formal or as a practical matter, and correction of the decree at this time would in no way compromise our continuing intention to effect a final allocation of the Lower Basin mainstream. Furthermore, awarding additional diversion rights to reflect the irrigable acreage not considered prior to the 1964 decree would correct a manifest injustice to the Tribes, who were not themselves before this Court in 1964, and it would do so with little, if any, prejudice to interests of other parties to this litigation.
A
The Court’s opinion excessively extols the principle of “finality,” but overlooks the caveat that “finality” means different things in different contexts, and that the law accords finality different weight depending on the context. First, the Court borrows support from formal, largely nondiscretionary doctrines such as res judicata. It admits, however, that res *644judicata has no applicability to this case, ante, at 619, for the simple reason that the omitted lands claims have been raised in the course of the same proceeding in which they were supposedly decided before, and that proceeding has not yet reached the stage of final judgment. In a case such as this, when a party seeks reconsideration of questions decided at an earlier stage of a single, continuing litigation, the law allows courts more discretion than in a case in which the party wants to upset a final judgment in another proceeding, before another judge. See generally IB J. Moore & T. Currier, Moore’s Federal Practice ¶¶0.401, 0.404[1J (1982) (hereinafter Moore); cf. United States v. United States Smelting Refining & Mining Co., 339 U. S. 186, 199 (1950).
A final judgment makes a difference. It marks a formal point at which considerations of economy, certainty, reliance, and comity take on more strength than they have before the judgment. A court’s decision to reconsider a prior ruling before the case becomes final, however, is ultimately a matter of “good sense.” Moore §0.404[10], at 573. Concern for finality remains an important policy, even before final judgment. In the absence of some overriding reason, a court should be reluctant to reopen that which has been decided merely to correct an error, even though it has the power to do so. See Messenger v. Anderson, 225 U. S. 436, 444-445 (1912). Nevertheless, federal courts have traditionally thought that correcting a manifest injustice was reason enough to reconsider a prior ruling, see Moore ¶0.404[1], p. 408, and, although they may hold a party to its failure to litigate a claim when it had the opportunity, they have regarded finality concerns as less compelling when the question at issue has never actually been contested, see Hartford Life Ins. Co. v. Blincoe, 255 U. S. 129, 136 (1921).1
*645The Court also uses “finality” in a more practical sense, appealing to the obvious benefits to society of having property rights be certain. This meaning of finality underlies the Court’s invocation of Abraham Lincoln and the development needs of the West. Ante, at 620-621. More importantly, it was central to the Court’s choice of an “irrigable acreage” standard in 1963, for that measure accorded the highest degree of certainty to all Lower Basin interests. Special Master Rifkind rejected Arizona’s proposal that the Indians be allocated only enough water to satisfy their presently foreseeable needs, precisely because that solution would be subject to re-evaluation in the future, “placing] all junior rights in jeopardy of the uncertain and the unknowable.” Rifkind Report 263-264. Therefore, he urged — and the Court held, 373 U. S., at 600-601:
“[T]he most feasible decree that could be adopted in this case, even accepting Arizona’s contention, would be to establish a water right for each of the five Reservations in the amount of water necessary to irrigate all of the practicably irrigable acreage on the Reservation. . . . This will preserve the full extent of the water rights created by the United States and will establish rights of fixed magnitude and priority so as to provide certainty for both the United States and non-Indian users.” Rifkind Report 265.
Thus, although the Court stresses Special Master Rifkind’s interest in a fixed and final decree, see ante, at 622-624, and n. 15, that interest is largely irrelevant to the question at hand. One can share Special Master Rifkind’s interest in having a fixed decree, and even Abraham Lincoln’s scorn for scoundrels in courthouse basements, and still think it desirable to correct the decree before it becomes fixed. Our interest in a fixed, reliable decree is well enough served if we make clear *646that it should not be subject to reopening, even to correct the kind of clear error that the Tribes and the United States have shown here, once this litigation becomes final.
The Court acknowledges that this litigation was far from final when the United States and the Tribes raised the claims now at issue, because the Court had not confirmed a list of the “present perfected rights,” or rights to use Colorado River mainstream flows that vested before the effective date of the Boulder Canyon Project Act of 1928, 43 U. S. C. § 617. Ante, at 611. The allocation system for the Lower Basin could not become final until an authoritative list of “present perfected rights” and their priority dates had been established.2 Article II of the 1964 decree identified a number of federal “present perfected rights,” including those of the Tribes, representing rights to divert about 900,000 acre-feet of mainstream flows per year. The 1964 decree, however, did not address any “present perfected rights” acquired under state law. The full list of “present perfected rights” was not submitted to or confirmed by this Court until 1979. See 439 U. S. 419. As quantified by our 1979 decree, state “present perfected rights” accounted for rights to divert well over 3 million acre-feet of mainstream flows. Thus, in 1977, *647when-the Tribes first sought to intervene in this litigation for the purpose of raising their omitted lands claims, and in 1978, when the United States moved for entry of a supplemental decree concerning the omitted lands, issues critical to the 1964 decree’s allocation system had yet to be finally determined.3
Furthermore, it has long been recognized that the primary object of this litigation was to establish a regimen for allocating the Lower Basin waters sufficiently reliable to permit Congress and Arizona to go forward with the Central Arizona Project, a massive public works effort to make Colorado River water available to agricultural interests in central Arizona. Tuttle Report 38-39; Meyers, The Colorado River, 19 Stan. L. Rev. 1, 73 (1966) (hereinafter Meyers). That purpose has been accomplished. The Central Arizona Project was authorized in 1968, and construction has now reached an advanced stage. But even at this late date the Project is still several years from completion. And until it is ready to begin diverting Colorado River water, the allocation system in our 1964 decree has little practical importance, because Arizona lacks the capacity to use most of the water rights allocated to it in the 1964 decree.
In sum, the interest in “finality” does not dispose of this case. Principles of judicial economy provide the sole basis for the Court’s refusal to correct the 1964 decree. But no significant adjudicative resources were expended on the omitted lands claims in the proceedings prior to the 1964 decree, because they were not raised at all. And, although the United States’ failure to identify the omitted irrigable lands 25 years ago should not be excused, I cannot join in depriving the Tribes permanently of significant rights to water on that basis alone, especially when I see little prejudice to the *648States from reopening the 1964 decree to the extent necessary to correct the error.4
B
The Tribes will suffer a manifest injustice if we fail to consider the omitted lands claims. Under the uncorrected 1964 decree, the Tribes stand to lose forever valuable rights to which they are entitled under the Court’s construction of the Executive Orders creating their reservations, 373 U. S., at 595-601. This loss occurs entirely because the United States failed to perform its obligations as trustee and advocate to present evidence to the Court of all irrigable lands within the reservations, or at least to make a record of its justification for not presenting such evidence.
It is certainly not the case that the United States made a considered decision to waive the Tribes’ claims to water for the omitted lands. Cf. ante, at 617-618, n. 7, and 622-623, n. 14 (suggesting otherwise). The existence of some omitted *649irrigable lands came to light at one point in the hearings, when an agricultural specialist mentioned that some mesa lands adjacent to irrigable acreage claimed by the United States could also be irrigated. App. to Brief for State Parties in Support of Exceptions 11. Special Master Rifkind immediately pressed the United States’ representative for an express waiver on the spot of all claims to water for irrigable acreage not identified in the pre-1964 hearings, but the attorney responded, “I am probably not authorized to give anything away that we ought to claim.” Id., at 12.5
Heckman v. United States, 224 U. S. 413 (1912), see ante, at 627, does not require us to make the Tribes bear the cost of the United States’ error. The relevant question in Heck-man, raised by non-Indian defendants, was whether individual Indians were necessary parties in a suit by the United States to set aside conveyances by those Indians of lands they were forbidden by statute to alienate, and over which the United States had significant trust responsibilities. 224 U. S., at 444. The Court held that the United States had power to enforce the statutory restrictions without the acquiescence of the Indians, and that by virtue of the restrictions the individual Indians had no interest in the subject matter of the suit. Id., at 445. In passing, the Court noted that representation of Indian interests by the United States “traces its source to the plenary control of Congress in legislating for the protection of the Indians under its care, and it recognizes no limitations that are inconsistent with the discharge of the national duty.” Ibid.
*650Were it not for the trust relationship recognized in Heckman and other cases, the United States’ litigation decisions could not estop the Tribes, who were not separately represented. Insofar as Heckman intimates that the United States’ power to-compromise Indian interests is not subject to judicial scrutiny, it has long since been repudiated by this Court. See, e. g., Shoshone Tribe v. United States, 299 U. S. 476 (1937); United States v. Creek Nation, 295 U. S. 103, 110 (1935); Cramer v. United States, 261 U. S. 219, 227-229 (1923). Instead, we have recognized that the United States’ relationship to Indian interests is much like that of a fiduciary to a beneficiary. Under the modern view, the “discharge of the national duty” requires sharp attention to the quality of the United States’ fulfillment of its trust obligations, including the obligation to represent Indian interests in litigation.
There has often been reason to question the quality of that representation, especially when rights to scarce water in the West were at stake. In 1973, the National Water Commission reported: “In the history of the United States Government’s treatment of Indian Tribes, its failure to protect Indian water rights for use on the Reservations it set aside for them is one of the sorrier chapters.” National Water Comm’n, Water Policies for the Future — Final Report to the President and to the Congress of the United States 475. President Nixon admitted as much in a 1970 message to Congress:
“The United States Government acts as a legal trustee for the land and water rights of American Indians. These rights are often of critical economic importance to the Indian people; frequently they are also the subject of extensive legal dispute. In many of these legal confrontations, the Federal government is faced with an inherent conflict of interest. The Secretary of the Interior and the Attorney General must at the same time advance both the national interest in the use of land and *651water rights and the private interests of Indians in land which the government holds as trustee.
. . There is considerable evidence that the Indians are the losers when such situations arise.” H. R. Doc. No. 91-363, pp. 9-10, 116 Cong. Rec. 23261 (emphasis in original).
The Court carefully explains that the United States had no “actual conflict of interest” with regard to Lower Basin water rights, by which it apparently means that the recognition of Indian water rights did not diminish other federally reserved water rights. See ante, at 627. I agree. Nevertheless, history discloses that the United States has not always taken such a narrow view of its interests in water rights controversies. On the Colorado River and elsewhere, it has constructed extensive water projects to serve nonfederal interests; congressional authorization of the Boulder Canyon Dam was the crucial event in the development of the Lower Basin, shaping this litigation from its inception. See 373 U. S., at 564-590. The United States has sometimes been slow to press Indian claims when they conflicted with those of politically influential non-Indian interests. See, e. g., Pyramid Lake Paiute Tribe v. Morton, 354 F. Supp. 252, 256-257 (DC 1973). See generally Federal Protection of Indian Resources: Hearings before the Subcommittee on Administrative Practice and Procedure of the Senate Committee on the Judiciary, 92d Cong., 1st Sess., 235-249, 907-914 (1971) (hereinafter Senate Hearings); F. Cohen, Handbook of Federal Indian Law 596-599 (1982) (hereinafter Cohen).6 We *652should not, therefore, leap to the conclusion that the irriga-bility of all reservation lands, including the omitted lands, “was fully and fairly litigated in 1963,” ante, at 628.
This case provides proof (if any is needed) that those with direct interests — economic, historical, spiritual — in the outcome of a case are their own best representatives. Upon entering this litigation, the Tribes swiftly exposed the extent of the United States’ pre-1964 neglect. I would not hold that the United States had so violated the ordinary standards of attorney care as to be liable for “inadequate” representation of the Indian interests in this litigation, if that were the standard of liability, on the basis of the mere fact that it failed to claim water rights for some irrigable acreage. But I do not find in this record any justification for the United States’ failure to present evidence on the omitted lands. Even if the United States did intend to waive the omitted lands claims, I see no good reason, before final judgment, to deny the Tribes a hearing on claims that have never been litigated. As a matter of justice, the Tribes deserve this chance to defend rights which should have been theirs.
C
In deciding whether to correct the 1964 decree, we should also consider any possible prejudice which the States might suffer as a result. Of course, the States would prefer that we not allocate additional water rights to the Tribes; at least at some point in the future, additional Indian rights may make the rights of junior state appropriators less certain. With regard to timeliness and finality, however, prejudice means prejudice from procedure rather than from the result. Hence, the important question is whether the States would *653be any worse off because the additional Indian rights were confirmed in 1983 rather than 1964.
The Special Master considered this issue at length and determined that the States would not be significantly prejudiced by adjustments in the 1964 decree. Tuttle Report 38-46.7 The whole question of reliance by the States, however, involves the highest degree of speculation. First, the amount of water entering the Lower Basin at Lee Ferry, Ariz., and available for use by Lower Basin interests has historically averaged far more than the 7.5 million acre-feet contemplated by the 1964 decree. See Rifkind Report 117. Until far more development occurs in the Upper Basin States, that situation can be expected to continue. Furthermore, improvements in irrigation, farming, and conservation technology may well permit more efficient exploitation of the present and future quantities of available water, so that more users will be accommodated by the same or less amounts of water.
In addition, the Tribes are not currently able to use all the rights allocated to them under the 1964 decree.8 Until sub*654stantial new irrigation systems or industrial plants are built, any additional water rights that the Tribes receive will have little or no practical effect on the availability of water to other Lower Basin interests. The Tribes can probably lease their rights to others with the consent of the United States, but they have not explored this option extensively. See Cohen 592-593; Meyers 71; cf. 2 Op. Solicitor of Dept, of Interior Relating to Indian Affairs 1917-1974, p. 1930 (1964). Even if the Tribes leased all of their rights to other Lower Basin users, it would merely mean that existing interests with the means to divert water from the Colorado River would pay a market rate for additional water. If the Tribes do not lease their rights, the water will simply be available for use by other Lower Basin interests, in accord with the allocation system established by the 1964 decree. In any event, non-Indian users will not be deprived of water in the near future on account of the rights at issue in this case.
In sum, correcting the 1964 decree to reflect additional irrigable acreage in the omitted lands would not harm the States more than they would have been harmed had the omitted lands been considered in framing the 1964 decree. In truth, Indian water rights are unlikely to affect state interests to any significant degree until well into the next generation, *655when all concerned will have had plenty of time to prepare. Yet if we foreclose the Tribes now from asserting their rights to water for the omitted lands, those rights will be lost forever, through no fault of their own. The balance of hardships in this case is decidedly in the Tribes’ favor. In order to avert a manifest injustice to the Tribes before this litigation becomes final and the allocation system in the 1964 decree begins to have a practical effect, I would reopen the 1964 decree to recognize additional water rights for the Tribes.
HH H-4
Reasonable judges might differ over some aspects of this case, but I would not have thought the Special Master’s solution to the boundary lands controversy was among them. The Court’s failure to approve a decree that includes a quantification of the water rights appurtenant to the disputed boundary areas serves no discernible purpose, and it is profoundly inconsistent with its emphasis in Part IV of its opinion on the ideals of finality, judicial economy, and predictability of water rights. At no point does the Court explain its rejection of the Special Master’s entirely reasonable proposal regarding the boundary lands.
In our 1963 opinion, we rejected Special Master Rifkind’s de novo determination of boundary disputes concerning two of the reservations, 373 U. S., at 601, and our 1964 decree was left open to the extent of permitting an award of additional water rights should the boundaries be “finally determined,” Art. 11(D)(5), 376 U. S., at 345. The 1979 decree recognized that the actual boundaries of all five reservations are subject to dispute. 439 U. S., at 421-422. At the outset of the current phase of this litigation, all parties agreed that it was time to bring the maximum degree of certainty possible to the Lower Basin allocation system, a task requiring “final determination” of the disputed boundaries, at least for the purpose of quantifying the Tribes’ entitlement to water. The United States and the Tribes urged before the Special Master that certain administrative determinations by *656the Secretary of the Interior had finally determined the boundaries of the reservations, where the disputed boundaries lay between reservation land and other federal lands.9 The States argued, as they had in 1963, that this Court should determine the relevant boundaries de novo.
The Special Master chose a middle course, calculated to put an end to further litigation in this Court. He took evidence on and determined the amount of irrigable acreage within the boundaries recognized by the Secretary of the Interior, and he calculated the corresponding water rights for inclusion in the final decree. However, he also recommended that the final decree include the following proviso:
“Provided, further, . . . that lands presently determined for this purpose to be within the boundaries of the above-named Reservations and later determined to be outside the boundaries of the above-named Reservations, as well as any accretions thereto to which the owners of such land may be entitled, should not be included as irrigable acreage within the Reservations and that the above specified diversion requirements of such land that is irrigable shall be reduced by the unit diversion quantities Usted in the [1979 decree].” Tuttle Report 282-283.
The effect of this proviso would be to grant the Indian Tribes the water rights appurtenant to the disputed boundary areas on a conditional basis. If the States succeeded in overturning any of the Secretary’s boundary determinations in an appropriate forum, the corresponding water rights — precisely quantified for each area in the Special Master’s Report, id., at 192-196, 239-277 — would automatically be subtracted from the Tribes’ entitlements.
*657The advantages of the Special Master’s proposal are obvious. First and foremost, it remains faithful to the approach taken in our 1963 opinion. On the one hand, it does not require this Court to decide in the first instance either what are the exact boundaries of the reservations or whether the Secretary’s administrative boundary determinations are binding on all parties for all purposes. On the other hand, it settles the maximum possible extent of Indian water rights. It allows the States to rely absolutely on that figure, and it informs them precisely how much water is at stake if they choose to litigate particular boundary questions in other forums. In 1963, the same considerations led us to adopt the “irrigable acreage” standard itself. Special Master Rifkind recommended rejecting an open-ended decree because it “would place all junior water rights in jeopardy of the uncertain and the unknowable,” Rifkind Report 264, whereas a fixed decree would “provide certainty for both the United States and non-Indian users,” id., at 265. Finally, the Special Master’s proposal would preclude further litigation in this Court over quantification of the water rights reserved for any boundary areas in fact within the reservations.
The Court disregards these virtues. Simply turning the clock back to 1964, it guarantees that the original jurisdiction litigation over Lower Basin water rights will proceed to another “round,” and possibly still more “rounds” thereafter, as one-by-one the border questions are settled by litigation. If any of the Secretary’s determinations are upheld, the Court will have to duplicate the efforts of the present Special Master. See ante, at 638.10 The full extent of the Tribes’ rights *658to divert mainstream water will remain uncertain for the near future, just as finality in this case begins to have practical importance. See supra, at 647.
For the reasons described in Part I-C, supra, awarding additional water rights to the Tribes works no immediate harm to state interests. The Court’s preference for prolonging this litigation and its attendant uncertainty is at odds with the principles upon which it resolves the omitted lands issue. I would accept the Special Master’s resolution of the boundary lands issue for purposes of framing a final decree in this action.
Ill
The Court’s disposition of the omitted lands and boundary lands issues makes it unnecessary for it to reach the remaining issues in this case. Although my own views would require us to reach those issues, I do not think it worthwhile to discuss them at any length. The States have filed a number of highly specific exceptions to the Special Master’s determinations regarding the irrigability of particular parcels. Although formal concepts of “plain error” and “abuse of discretion” do not apply to the recommendations of special masters in original jurisdiction litigation, the care with which the present Special Master has explained his conclusions on these technical issues demands respect, and I would overrule the States’ exceptions. The United States has also filed four exceptions. The first asks that we recognize for purposes of our decree the Secretary of the Interior’s resolution of an *659additional border question concerning the Fort Mojave Reservation; the others involve essentially clerical matters of conforming the Special Master’s recommended decree to our two prior decrees. I would sustain the exceptions of the United States.
The equity doctrine of “changed circumstances,” see ante, at 624-625, reflects many of the same principles. Yet even if changed circumstances are necessary to modify an injunction — and I doubt that an equity court would turn its back on manifest injustice — they have never been the sine *645qua non of adjusting a decree in the process of making it final. The question before us is whether we should do that.
It is unnecessary to describe fully the complex structure of our 1964 decree. Suffice it to say that the Indian Tribes’ rights at issue in this case are among the “present perfected rights,” but they are not the only such rights. These rights are important because the Secretary of the Interior has an obligation to satisfy them to their full extent, and that water is charged against the States’ overall entitlements under the 1964 decree. Furthermore, in drought years “present perfected rights” cannot be made to bear pro rata reductions along with other water users; rather, the Secretary is obligated to satisfy them in full, starting with the right established first in time and proceeding chronologically (except for the Indian rights, which must be satisfied first regardless of priority, 439 U. S. 419, 421 (1979)). As a practical matter, then, the more “present perfected rights” there are, the less certain it is that other users will receive a specific amount of water in any given year, especially in years when mainstream flows are less than the 7.5 million acre-feet benchmark used in the 1964 decree.
The 1979 decree was handed down before we acted on the Tribes’ motions to intervene or any of the claims now before the Court. The decree expressly left these matters open for resolution and referred them to Judge Tuttle as Special Master. Id., at 421-422, 436-437.
The Court suggests that if we reopened the question of irrigable acreage we would also have to reconsider the “irrigable acreage” standard itself. See ante, at 626-626. In raising that specter, the Court ignores the obvious distinction between the standard and its application to the omitted lands. No issue was the subject of more controversy in the proceedings leading up to our 1964 decree than the “irrigable acreage” standard. Unlike the actual quantification of the acreage, the standard was discussed extensively, both in Special Master Rifkind’s report, at 257-266, and in the Court’s opinion, 373 U. S. 546, 600-601 (1963). The “irrigable acreage” standard has been fully and fairly litigated. Nor does the Court’s opinion or Special Master Rifkind’s report indicate that some other standard of measurement would have been chosen had the Court been apprised of the irrigable acreage in the omitted lands. This Court adopted the “irrigable acreage” standard for the reasons stated in its opinion — it is the only “feasible and fair way by which reserved water for the reservations can be measured,” id., at 601. It reflects the purposes for which the reservations were created, and once final it need not be readjusted in light of changed circumstances, unlike an equitable measure linked to current or expected population. If a few acres worth of water more or less would have changed our decision, we would not have rejected the argument that Indian water rights be determined by familiar equitable principles rather than by the more objective standard.
See ante, at 622-623, n. 14. A close reading of the exchange between Special Master Rifkind and the Government attorney reveals that the Special Master did not continue to press his demand for a binding waiver. In light of the United States’ delicate trust responsibilities in Indian water cases, it would have been improper to require the attorney to make a split-second decision to concede an important class of claims in response to surprise testimony from a witness. Cf. Manual for Complex Litigation § 1.80, p. 89 (1982); Chayes, The Role of the Judge in Public Law Litigation, 89 Harv. L. Rev. 1281, 1291, 1297-1298 (1976).
There are many ways of compromising a claim besides making a decision not to press it. Devoting fewer resources to investigating and preparing the claim than its economic importance would warrant has the same effect. In eases such as this, the Justice Department is responsible for pressing the Indians’ claims in court, but the Interior Department and the experts it employs are responsible for developing the facts of the claim and bringing it to the attention of the Justice Department. The practical result of this bifurcated responsibility may often be to confer effective *652power to waive Indian claims on Interior Department hydrologists and agricultural experts. See Senate Hearings 445-449 (testimony of W. Kiechel, Jr., Deputy Assistant Attorney General, Land and Natural Resources Division).
The Special Master observed that in 1968 Congress authorized construction of the Central Arizona Project based on projections of mainstream flows available for diversion by Arizona far lower than current projections, so that it is not possible to argue that the Central Arizona Project would not be commercially viable if the Indians receive additional water rights. Tuttle Report 38-41; see S. Rep. No. 408, 90th Cong., 1st Sess., 18-21, 32-35 (1967). The Special Master also found that water would be available to meet the full diversion capacity of water projects begun by Nevada after 1964. Tuttle Report 44-46. The Metropolitan Water District of Southern California — the junior major appropriator in California— presented some evidence of reliance, but did not “fully explain why [it] will receive less water if the Tribes receive additional water rights.” Id., at 42. In any event, under current projections of demand the Metropolitan Water District will not be ready to use its existing entitlements before the year 2010. Ibid.
From 1975, when the Fort Mojave Tribe began to use its water for the first time, through 1981, the Tribes collectively diverted only 77% of the water to which they were entitled under the 1964 decree. In individual *654years, diversions ranged from 83% of the 1964 decree awards (1981) to 72% (1978). U. S. Dept. of Interior, Bureau of Reclamation, Compilation of Records in Accordance with Article V of the Decree of the Supreme Court of the United States in Arizona v. California Dated March 9,1964 — Calendar Years 1925-1981. The Chemehuevi diverted no water at all, although they are entitled to 11,340 acre-feet a year, ibid., because there appears to be no diversion system in place on their reservation, either for purposes of irrigation or for other development. See Senate Hearings 1075 (testimony of R. Esquerra). The Special Master’s Report makes clear that substantial capital investment would be required before the Tribes could begin to use additional water. See, e. g., Tuttle Report 165-184,242-248. On the Fort Mojave Reservation alone the United States’ expert estimated that over $2.1 million would be required to develop six units of land for which the United States claimed additional water rights. See United States Exhibits 132-140.
The Court determines that other disputed boundaries have been “finally determined” by judicial adjudications that the States have not challenged. It approves amending the 1964 decree to include water rights appurtenant to these parcels. Ante, at 640-641. To this extent, I concur in Part V of the Court’s opinion.
The Court seems to believe that pending litigation in the Southern District of California involving only some of the boundary issues presented by this case, as well as only some of the parties, provides an appropriate forum for resolving the boundary disputes once and for all. Ante, at 639. It suggests that other parties enter the lawsuit voluntarily, and that they use it to decide additional issues. Ante, at 639, n. 29. However, under the Court’s ruling today the States have absolutely no reason to prosecute additional claims — as long as the boundary issues are not decided, the water *658rights that turn on them belong to the States. (As defendant, of course, the United States has no choice but to litigate.) The Court also makes the unprecedented suggestion that we might be willing to decide the boundary questions de novo if the States’ District Court suit is barred by lack of standing, sovereign immunity, or the statute of limitations. Ante, at 638. I would not leave that impression. Because “[c]ertainty of rights is particularly important with respect to water rights in the Western United States,” ante, at 620, such results in the District Court would “finally determine” the boundaries of the reservations within the meaning of Article 11(D)(5) of the 1964 decree.