OPINION
WILKINSON, Circuit Judge;In this case an insurer comes to federal court seeking a declaratory judgment regarding its obligation to defend and indemnify its insured when the underlying mat*236ter for which the insured claims coverage is the subject of an action already pending in state court. The issue of whether a federal court should decline to entertain an insurer’s declaratory action when it is closely related to pending state court litigation has been one of recurrent difficulty. In the present version, Vines, a tenant of appellant Harris, brought non-removable, purely state law claims in Maryland state court against Harris alleging personal injuries stemming from lead poisoning. Harris’ liability insurer at Lloyd’s, London then brought this related, also purely state law action in federal court seeking a declaration that it owed no duty to defend or indemnify Harris on Vines’ claims. After declining to dismiss the case, the district court awarded summary judgment to the insurer.
We reverse. The questions of state law are close ones, as illustrated by the disagreement between the district court and the dissenting opinion over the insurer’s duty to defend. When considered in light of these problematic state law issues, the relevant state interests, in this case should have led the district court to exercise its statutorily provided discretion to decline to entertain this declaratory judgment action.
I.
Appellant Nathan Harris owned a house at 509 North Gilmor Street, Baltimore, Maryland, which he rented to Elizabeth Davis. On March 19, 1981, Harris procured a liability insurance policy from a group of underwriters at Lloyd’s, London covering the house for a period of one year. Harris had not had liability insurance for this property prior to that time. The policy promised to indemnify Harris for all damages for bodily injury and property loss arising out of the ownership, maintenance, or use of the insured premises during the policy period. The policy defined a covered “occurrence” as “an accident, including continuous or repeated exposure to conditions, which results in bodily injury or property damage neither expected nor intended from the standpoint of the insured.” Lloyd’s also promised to defend Harris in any action seeking such damages.
Elizabeth Davis was the legal guardian for Charlese Vines, who began living with Davis in the house shortly after Vines’ birth in 1978. Vines tested positive for elevated blood lead levels first in March 1980 and again at several points in 1980 and 1981. Lead paint violation notices regarding the house were issued by the Baltimore City Health Department to Harris in March and June of 1980, but abatement did not occur until June of 1981.
On February 13, 1989, Vines and Davis both filed suit against Harris in the Circuit Court for the City of Baltimore. Based solely on Maryland law, the complaint alleged that on or about August 1980, Vines had become seriously and permanently injured as a result of exposure to the lead paint in the house. After being notified of this claim, Lloyd’s provided Harris with a defense under a reservation of rights. This action is still pending in state court.
After the action had been filed in the Maryland court, appellee George Mitche-son, acting for and on behalf of certain other underwriters at Lloyd’s, filed this declaratory action in the federal district court for the District of Maryland on June 9, 1989. Lloyd’s sought a declaration that it was not obligated to defend or indemnify Harris regarding the claims made by Vines and Davis. Because Vines’ lead poisoning was first discovered prior to the inception of coverage, Lloyd’s alleged that the harm to Vines constituted an occurrence that was not covered by the policy. See Harford Mut. Ins. Co. v. Jacobson, 73 Md.App. 670, 536 A.2d 120, 127 (Md.Ct.Spec.App.1988). Lloyd’s also contended that Harris’ failure to advise it that the property had been cited for lead paint violations voided his coverage and that the alleged injuries fell within the policy’s pollution exclusion. In response to these contentions, Harris argued that the time of the first occurrence should be judged by when the injuries were first diagnosed, Jacobson, 536 A.2d at 127, and that there had been a second occurrence during the policy’s coverage period. Harris also argued under *237Jacobson that Lloyd’s had, at a minimum, a duty to defend him because the uncertainties over the circumstances surrounding the lead poisoning were “enough to indicate a potentiality that the injury in question occurred during the time the policy was in effect.” Id. 536 A.2d at 123; see also Brohawn v. Transamerica Ins. Co., 276 Md. 396, 347 A.2d 842, 850 (1975).
Lloyd’s moved for summary judgment on May 17, 1990. Harris opposed the motion and moved to dismiss the lawsuit. Focusing primarily upon the operative term “occurrence,” the district court held that Maryland law did not obligate Lloyd’s either to defend or indemnify Harris and accordingly granted summary judgment to the insurer. Harris now appeals those rulings.
II.
We acknowledge at the outset that the district court did possess diversity jurisdiction under 28 U.S.C. § 1332(a) to entertain this declaratory judgment action. The critical question, however, is whether the district court should have exercised the jurisdiction it possessed. Under the Declaratory Judgment Act, federal courts have discretion in deciding whether to hear a declaratory action. 28 U.S.C. § 2201 (“In a case of actual controversy within its jurisdiction, ... any court of the United States, upon the filing of an appropriate pleading, may declare the rights and other legal relations, of any interested party seeking such declaration, whether or not further relief is or could be sought.” (emphasis added)); see also A.L. Mechling Barge Lines, Inc. v. United States, 368 U.S. 324, 331, 82 S.Ct. 337, 341, 7 L.Ed.2d 317 (1961); 10A Charles A. Wright et al., Federal Practice and Procedure § 2759 (2d ed.1983). In reviewing a district court’s decision on whether to entertain such an action, an appellate court does not approach the case in a wholly deferential posture. Rather, as numerous circuits have agreed, the appellate court must exercise its own judgment in reviewing the various interests at stake. See, e.g., Cincinnati Ins. Co. v. Holbrook, 867 F.2d 1330, 1333 (11th Cir.1989) (per curiam); Fireman’s Fund Ins. Co. v. Ignacio, 860 F.2d 353, 354 (9th Cir.1988) (per curiam); International Harvester Co. v. Deere & Co., 623 F.2d 1207, 1217 (7th Cir.1980); Beacon Constr. Co. v. Mateo Elec. Co., 521 F.2d 392, 397 (2d Cir.1975); 6A James W. Moore et al., Moore’s Federal Practice ¶ 57.08[2] (2d ed. 1991).
In our review of the district court’s decision to entertain this declaratory action, we focus on two significant state interests that weigh in favor of dismissal.
A.
The first concern supporting dismissal of the declaratory action is the state’s interest in deciding questions of state law — an interest that not surprisingly has its jurisprudential roots in Erie R.R. Co. v. Tompkins, 304 U.S. 64, 58 S.Ct. 817, 82 L.Ed. 1188 (1938). As was well established by Erie and its progeny, state law ordinarily provides the rule of decision in pure diversity cases. There exists an interest in having the most authoritative voice speak on the meaning of applicable law, and that voice belongs to the state courts when state law controls the resolution of the case.
Obviously, the interest in having the law of a given sovereign determined by the courts of that sovereign is not absolute; there are countless examples of federal courts interpreting state law and vice ver-sa. In conferring federal jurisdiction under Article III, Congress can plainly supersede the state interest in having state courts interpret state law. In the ordinary diversity case, for example, there are paramount federal interests in having federal courts resolve questions of state law and accord protection to out-of-state parties.In declaratory actions, however, Congress has diminished the strength of these interests by making the exercise of such diversity jurisdiction discretionary with the court. The obligation of the federal judiciary to resolve all questions within its jurisdiction absent special circumstances, see Colorado River Water Conservation Dist. v. United States, 424 U.S. 800, 817-18, 96 S.Ct. 1236, 1246, 47 L.Ed.2d 483 (1976), must thus be qualified in a declaratory action. Indeed, *238the Third Circuit has stated that “by the terms of the Declaratory Judgment Act itself,- a fundamental reason for the exceptional nature of a Colorado River stay— the ‘virtually unflagging obligation of the federal courts to exercise the jurisdiction given them/ — is not present in declaratory judgment cases.” Terra Nova Ins. Co. v. 900 Bar, Inc., 887 F.2d 1213, 1222 (3d Cir.1989) (quoting Colorado River, 424 U.S. at 817, 96 S.Ct. at 1246); see also Mission Ins. Co. v. Puritan Fashions Corp., 706 F.2d 599, 601 n. 1 (5th Cir.1983) (statute, not exceptional circumstances test, controls exercise of discretion in declaratory actions); Allstate Ins. Co. v. Best, 728 F.Supp. 1263, 1269 (D.S.C.1990) (in deciding questions of state law in declaratory action, there exists “no federal interest whatsoever in their ultimate resolution”). In other words, in declaratory actions Congress has afforded the federal courts a freedom not present in ordinary diversity suits to consider the state interest in having state courts determine questions of state law.
In another area involving the discretionary power of federal courts to hear cases, the Supreme Court has recognized the desirability of having state courts interpret questions of state law. In United Mine Workers v. Gibbs, 383 U.S. 715, 86 S.Ct. 1130, 16 L.Ed.2d 218 (1966), the Court discussed the situation where a federal court obtains jurisdiction over a state law claim via pendent jurisdiction only to have the federal claim dismissed before trial. While recognizing that the federal court certainly has jurisdiction to- hear such a case, the Court stated that the federal court should frequently exercise its discretion to dismiss the case in order to avoid “[njeedless decisions of state law” when a state court could provide “a surer-footed reading of applicable law.” Id. at 726, 86 S.Ct. at 1139. The Court underscored that the federal interests supporting federal resolution of pendent state claims recede when the federal questions are dismissed, see id.; the state interest in having state courts resolve the remaining questions of state law rises correspondingly.1 Much the same state interest in having state courts resolve difficult questions of state law is recognized in the practice of certification, see Lehman Bros. v. Schein, 416 U.S. 386, 391, 94 S.Ct. 1741, 1744, 40 L.Ed.2d 215 (1974), and in the various forms of federal abstention. See, e.g., Railroad Comm’n v. Pullman Co., 312 U.S. 496, 61 S.Ct. 643, 85 L.Ed. 971 (1941).
The state interest in deciding the issues in the declaratory action is particularly strong in this case. All questions in the declaratory action depend solely upon state law. Cf. Mraz v. Canadian Universal Ins. Co., 804 F.2d 1325 (4th Cir.1986) (entertaining declaratory action regarding duty to defend and indemnify which involved interpretation of federal law, the Comprehensive Environmental Response, Compensation, and Liability Act of 1980, 42 U.S.C. § 9601 et seq.). The primary litigation, which seeks recovery from Harris for the injuries surrounding Vines’ lead poisoning, was pending in state court at the time this action was filed and involves only non-removable questions of state law. Absent a strong countervailing federal interest, the federal court here should not elbow its way into this controversy to render what may be an “uncertain and ephemeral” interpretation of state law. Pennhurst State School & Hosp. v. Halderman, 465 U.S. 89, 122 n. 32, 104 S.Ct. 900, 920 n. 32, 79 L.Ed.2d 67 (1984). Lloyd’s concedes that the only federal interest at issue in this case is the same interest federal courts have in deciding any diversity case. Given the previously filed state litigation on the primary claims and the exclusively state law nature of the declaratory action, however, the state interest in resolving matters of state law should have guided the exercise of federal judicial discretion.
*239B.
The second state interest favoring dismissal of the declaratory action is that of resolving all litigation stemming from a single controversy in a single court system. While the aforementioned interest of a state in resolving for itself questions of state law is rooted in the philosophy of judicial federalism, this second interest is grounded in the equally important pragmatic concerns of efficiency and comity.
In deciding whether to entertain a declaratory judgment action, a federal court should analyze whether its resolution of the declaratory action will settle all aspects of the legal controversy. This court has long recognized that it makes no sense as a matter of judicial economy for a federal court to entertain a declaratory action when the result would be to “try a controversy by piecemeal, or to try particular issues without settling the entire controversy.” Aetna Cas. & Sur. Co. v. Quarles, 92 F.2d 321, 325 (4th Cir.1937); accord 10A Charles A. Wright et al., supra, § 2759. For the federal court to charge headlong into the middle of a controversy already the subject of state court litigation risks “[gratuitous interference with the orderly and comprehensive disposition of [the] state court litigation.” Brillhart v. Excess Ins. Co., 316 U.S. 491, 495, 62 S.Ct. 1173, 1176, 86 L.Ed. 1620 (1942).
It hardly husbands scarce judicial resources to allow separate suits stemming from the same overall controversy and involving overlapping issues to proceed simultaneously on parallel tracks. The inefficiencies that could result are not hard to imagine. For example, if a federal court were to render a declaration that the insurer had no duty to defend, the ongoing state litigation could be delayed considerably due to time consuming substitution of counsel. Similarly, a federal declaration that an insurer had no duty to indemnify could be rendered totally unnecessary by a subsequent state verdict for the insured in the underlying state action. See Indemnity Ins. Co. v. Schriefer, 142 F.2d 851, 853 (4th Cir.1944) (approving dismissal of declaratory action because duty to indemnify could be decided as well after the state court litigation). Such potential problems reinforce the inherent appeal of having all litigation stemming from a single controversy resolved in a single court system. See generally Gibbs, 383 U.S. at 724-25, 86 S.Ct. at 1138. While inefficiencies can of course occur within a single court system, the prospects for coordinated management and alleviation of abrasion are greater when the litigation is handled under one jurisdictional roof. Because the principal litigation stemming from the allegations of lead poisoning is non-removable, the Maryland courts are the only system with the jurisdictional power to resolve all facets of this controversy, and they are also in the best position to assess how best to handle any satellite suits.
Running alongside these basic efficiency concerns is the interest in promoting comity between federal and state courts. An important element in fostering that spirit of cooperation and respect must be a reluctance on the part of federal courts to entertain a declaratory action that could result in entanglement between the two court systems. In many declaratory actions brought to resolve a duty to defend or indemnify an insured in a controversy already the subject of state court litigation, there will be overlapping issues of fact or law between the state and federal actions. Here, for example, the district court stated that it would have to decide a critical issue that “will be addressed in the pending state court action” — i.e., when Vines’ lead poisoning took place. The insured may well be collaterally estopped from relitigating the overlapping issues decided in the federal action. See Zurich Ins. Co. v. Alvarez, 669 F.Supp. 307, 309 (C.D.Cal.1987); Allstate Ins. Co. v. Harris, 445 F.Supp. 847, 851 (N.D.Cal.1978); Restatement (Second) of Judgments § 87 cmt. a (1982). Such issue preclusion will likely “frustrate the orderly progress” of state court proceedings by leaving the state court with some parts of a case foreclosed from further examination but still other parts in need of full scale resolution. Phoenix Ins. Co. v. Harby Marina, Inc., 294 F.Supp. 663, 664 *240(N.D.Fla.1969). Additionally, the state court will likely have to consult federal law to ascertain the preclusive principles at work, see Harnett v. Billman, 800 F.2d 1308, 1312-13 (4th Cir.1986), thereby creating further entanglement. While the parties dispute whether Maryland courts would entertain a declaratory action under circumstances such as these, see Allstate Ins. Co. v. Atwood, 319 Md. 247, 572 A.2d 154, 157-58 (1990); Brohawn, 347 A.2d at 849, it is ultimately for Maryland courts to decide both whether questions of Maryland law should be entertained in a declaratory action instead of being reserved for the primary litigation and, if such a declaratory action is heard, whether it should be resolved before the primary suit. A system of judicial federalism has enough inherent friction without the added aggravation of unnecessary federal declarations on questions such as those at issue here.
C.
The aforementioned interests assume greater saliency where, as here, the issues of state law are close. Simply by way of illustration of the difficulty, we note that the district court and the dissenting opinion come to opposite conclusions on whether Lloyd’s has a duty to defend Harris in the state court litigation. Our dissenting brother focuses on the Maryland Court of Appeals’ statement that “[e]ven if a tort plaintiff does not allege facts which clearly bring the claim within or without the policy coverage, the insurer still must defend if there is a potentiality that the claim could be covered by the policy.” Brohawn, 347 A.2d at 850 (emphasis in original). Because our dissenting colleague believes that Harris established a potentiality of coverage, he would reverse the district court’s declaration that Lloyd’s had no duty to defend.
While the district court noted the Bro-hawn language relied upon by the dissent, it looked to a later statement by the Maryland Court of Appeals that “when the question of coverage or defenses under the language or requirements of the insurance policy is separate and distinct from the issues involved in the tort suit, the ‘potentiality rule’ ... has no application.” St. Paul Fire & Marine Ins. Co. v. Pryseski, 292 Md. 187, 438 A.2d 282, 286 (1981). The district court also noted that under Maryland law the “duty to defend continues, if actual or potential coverage is established, until such coverage is negated.” 7416 Baltimore Ave. Corp. v. Penn-America Ins. Co., 83 Md.App. 692, 577 A.2d 398, 401 (Md.Ct.Spec.App.1990). Having decided that Harris was not entitled to coverage because no “occurrence” took place within the policy period, the district court concluded that any duty to defend no longer existed.
We do not seek to resolve the dispute between the district court and our dissenting colleague on the question of the insurer’s duty to defend. Indeed, the inadvisability of seeking to resolve it is underscored not only by the closeness of the state law question, but also by the fact that, as to the duty to indemnify, the dissent “agree[s] generally that deciding the issue of coverage would unnecessarily entangle the federal action with the state proceeding.”2
What this circuit once said in the context of a federal declaratory action on an insurer’s duty to indemnify is equally applicable here:
To have sustained the suit for declaratory judgment would have been to drag this essentially local litigation into the *241federal courts and to defeat the jurisdiction of the state courts over it merely because one of the parties to the litigation happened to have indemnity insurance in a foreign insurance company.
Schriefer, 142 F.2d at 853.
III.
For the foregoing reasons, the judgment of the district court is reversed and the case is remanded with directions to dismiss it. The dismissal, of course, is to be without prejudice to any suit that may be brought in state court.
REVERSED AND REMANDED.
. We recognize, of course, that the pendent jurisdiction context differs from the declaratory judgment situation in that the federal courts have an independent diversity basis for exercising jurisdiction over the state claim in the latter case. This does not weaken the analogy, however, because in both instances, the critical factor to be considered is the discretion with which the federal court is vested to decide .whether to entertain the state law claim.
. Our conclusion is not altered by the fact that Lloyd’s alleged in its complaint (1) that the defendant, when applying for his policy, failed to inform Lloyd's that the property in question had been cited for lead paint violations and (2) that the policy contained a pollution exclusion which prohibited coverage in this case. The claim relating to the pollution exclusion was apparently abandoned by Lloyd’s following discovery. Additionally, the decision of the district court did not rest primarily on the absence of notice or the presence of an exclusion but instead upon that court’s determination of when an "occurrence," as defined by the operative insurance contract, would be deemed by Maryland courts to have taken place in a toxic exposure case. While the Maryland courts might well be persuaded that the district court's view of the matter is the correct one, we think they should be given the chance to conclude otherwise.