OPINION OF THE COURT ON PETITION FOR EXTRAORDINARY RELIEF IN THE NATURE OF A WRIT OF PROHIBITION
CAIRNS, Senior Judge:In a Petition for Extraordinary Relief in the Nature of a Writ of Prohibition, petition*592er asks this court to dismiss charges that were referred to a special court-martial convened in the Republic of Korea. The issue is whether court-martial jurisdiction was severed when the petitioner was discharged after arraignment but before charges were resolved by lawful authority. We have considered the record on the motion, the petitioner’s pleadings,1 the government’s answer to show cause, and oral argument. We hold that in personam jurisdiction over the petitioner was lost upon his discharge from the service, and we therefore grant the petition.
I. FACTS
The petitioner was assigned to Headquarters and Headquarters Company, Eighth United States Army (EUSA), Korea. He was then attached by written order to Headquarters and Headquarters Company, 2d Infantry Division (ID), for the administration of military justice and other purposes. He performed his military duties in the 2d ID area of responsibility, but served under an EUSA chain of supervision comprised primarily of civilian employees. Petitioner’s attachment orders to 2d ID lapsed in 1994, but he continued to perform the same duties under the same supervisory chain.
On 14 May 1996, the Commanding General, 2d ID, referred charges to a special court-martial empowered to adjudge a bad-conduct discharge. The petitioner was arraigned on 30 May 1996. During an Article 39(a), UCMJ, 10 U.S.C.A. § 839(a) session held on 5 June 1996, the case was set for trial on 26 June 1996.
Concurrent with the processing of court-martial charges by the 2d ID, the EUSA Transition Center issued orders on 20 May 1996, directing petitioner to complete final outproeessing for discharge from the United States Army. The discharge was to be effective on 15 June 1996, the date of petitioner’s expiration of term of service (ETS). Petitioner’s civilian supervisor, who was concerned that the pending discharge and eourtmartial were inconsistent actions, delivered a copy of petitioner’s ETS outproeessing orders to the office of the commander who preferred the court-martial charges. Nevertheless, the convening command neither flagged petitioner’s records nor undertook any other action to halt the discharge that was being processed by EUSA, the petitioner’s unit of assignment.
The petitioner proceeded to complete his outproeessing in accordance with the normal EUSA requirements for all soldiers being discharged upon their ETS. Consistent with standard procedures, his final pay was computed and arrangements were made for him to receive his final pay by mail. The petitioner’s unconditional ETS discharge was approved by proper authority, and his discharge certificate was delivered to him on 14 June 1996. The next day, the petitioner flew to his home of record in the United States, using an airplane ticket provided by the Army.
The record reveals that the petitioner’s EUSA and 2d ID chains of command and supervision did not have a clear understanding of petitioner’s assignment status or their respective responsibility for him. Consequently, neither the discharge action nor the court-martial action was effectively coordinated between the two commands.
On 24 June 1996, the court-martial reconvened in the absence of the accused. The trial defense counsel moved to dismiss all charges for lack of jurisdiction. The military judge denied the motion, found facts which suggest that the discharge was induced by mistake, or perhaps fraud, and ruled “as a matter of law” that:
Once charges were preferred and court-martial jurisdiction attached, the accused was not eligible to be discharged until action on the charges (trial, sentence, and appellate review or dismissal of the charges) was completed by lawful authority. Actions taken to the contrary were without actual authority and do not termi*593nate jurisdiction.2
II. LAW
In the landmark case of United States ex rel. Toth v. Quarles, 350 U.S. 11, 76 S.Ct. 1, 100 L.Ed. 8 (1955), the Supreme Court found that Article I, Section 8, Clause 14 of the Constitution — -which grants Congress the power to “make Rules for the Government and Regulation of the land and naval Forces” — established the outer limit of court-martial jurisdiction over discharged soldiers: A soldier who violates military law while a member of the Army, but who is discharged prior to any action being taken with a view toward court-martial, is a civil person and may not be subjected to court-martial jurisdiction. In cpnstruing Article I to restrict court-martial jurisdiction to persons who are actually members or part of the armed forces, the Supreme Court held that Congress could not rely on the notion of continuing jurisdiction to extend court-martial jurisdiction to “civilian ex-soldiers who [have] severed all relationship with the military and its institutions.” Toth, 350 U.S. at 14, 76 S.Ct. at 4. The Court reasoned that the relationship between the Army and those soldiers who have been unconditionally discharged is so attenuated that the exercise of court-martial jurisdiction exceeds “the least possible power adequate to the end proposed,” that is, the regulation of the land and naval forces. See Toth, 350 U.S. at 23, 76 S.Ct. at 8.
The Constitution is not the only limit on court-martial jurisdiction. Inasmuch as court-martial jurisdiction is created by statute, jurisdiction is also limited to those discharged persons who are constitutionally brought under it through the jurisdictional provisions of the Uniform Code of Military Justice [hereinafter UCMJ]. See United States v. Smith, 4 M.J. 265, 266 (C.M.A. 1978); Rule for Courts-Martial 202(a) [hereinafter R.C.M.]. In construing the scope of congressional grants of jurisdiction, the Supreme Court ensures that Congress acts on “the implicit assumption that without a grant of congressional authority military courts [are] without power to try discharged or dismissed soldiers for any offenses committed while in the service.” United States ex rel. Hirshberg v. Cooke, 336 U.S. 210, 215, 69 S.Ct. 530, 533, 93 L.Ed. 621 (1949). The courts have also recognized, however, that congressional grants of court-martial jurisdiction may, under limited circumstances, include the power to continue the exercise of that jurisdiction after a person has lost statutory jurisdictional status through separation from the service. Thus far, courts have applied the principle of continuing attached jurisdiction sparingly. Continuing jurisdiction has been applied to sustain jurisdiction only over those already tried and convicted while in a status subject to the code. Carter v. McClaughry, 183 U.S. 365, 22 S.Ct. 181, 46 L.Ed. 236 (1902); United States v. Speller, 8 U.S.C.M.A. 363, 24 C.M.R. 173, 1957 WL 4734 (1957). The Courts in McClaughry and Speller applied continuing jurisdiction for the limited purpose of completing convening authority action, appellate review, and sentence execution.
The Court of Military Appeals (now the United States Court of Appeals for the Armed Forces) has recognized that, in the absence of statutory authorization, a valid unconditional discharge will sever court-martial jurisdiction. That court has held that “[i]t is black letter law that in personam jurisdiction over a military person is lost upon his discharge from the service, absent some saving circumstance or statutory authorization.” United States v. Howard, 20 M.J. 353, 354 (C.M.A.1985). When no explicit statutory authority over discharged soldiers exists, the courts have recognized “saving circumstances” in only very narrowly drawn situations. Although not specifically recognizing the concept of continuing attached jurisdiction as a “saving circumstance,” the courts have sustained jurisdiction over soldiers who are tried and sentenced by court-martial while subject to the code, but discharged prior to final action. See Carter, 183 U.S. 365, 22 S.Ct. 181; Speller, 8 U.S.C.M.A. *594363, 24 C.M.R. 173. Otherwise, military courts have found “saving circumstances” sufficient to sustain court-martial jurisdiction over discharged soldiers only where the discharge is either void or conditional. See e.g., United States v. Garvin, 26 M.J. 194 (C.M.A. 1988) (discharge barred by statute); United States v. Johnson, 6 U.S.C.M.A. 320, 20 C.M.R. 36, 1955 WL 3535 (1955) (re-enlistment discharge without intent to release unconditionally); see also United States v. Banner, 22 C.M.R. 510, 516, n. 1, 1956 WL 4820 (A.B.R.1956) (general discussion of validity of discharges).
III. DISCUSSION
Looking to the plain language of Articles 2 and 3, UCMJ, the jurisdictional provisions of the UCMJ, it is clear that Congress has not explicitly extended court-martial jurisdiction over discharged persons situated like the petitioner. Article 2, UCMJ, lists classes of persons subject to the code. Article 3, UCMJ, describes certain discharged persons whose status as members of the armed forces has apparently ended, but who are nevertheless made subject to the code by statute. Article 2(a)(1), UCMJ, comes the closest to touching the petitioner’s situation: it establishes jurisdiction over members of the armed forces “awaiting discharge after expiration of their terms of enlistment.” However, because that language is explicitly limited to “those awaiting discharge,” it can hardly be construed to include one who has been discharged. Since no explicit provision of the statute authorizes jurisdiction over the petitioner, we must ascertain whether some “saving circumstance” exists sufficient to sustain court-martial jurisdiction.
A. Continuing Attached Jurisdiction
The government argues that because the petitioner was arraigned prior to being discharged, jurisdiction attached while petitioner was subject to the code. They concluded that such attached jurisdiction continues until the charges are resolved, even though a valid discharge was issued prior to trial. They cite United States v. Speller, 8 U.S.C.M.A. 363, 24 C.M.R. 173, 1957 WL 4734 (1957). See also United States v. Engle, 28 M.J. 299 (C.M.A.1989); United States v. Montesinos, 28 M.J. 38 (C.M.A.1989); United States v. Jackson, 3 M.J. 153 (C.M.A. 1977); United States v. Entner, 15 U.S.C.M.A. 564, 36 C.M.R. 62, 1965 WL 4785 (1965); United States v. Sippel, 4 U.S.C.M.A. 50,15 C.M.R. 50,1954 WL 2251 (1954).
Speller held that a valid discharge issued after court-martial findings and sentencing does not cut off authority to complete post-trial action and appellate review. Speller is consistent with precedent that recognizes the authority of the military to continue its review and appeal processes and to execute sentences adjudged by courts-martial over persons who were tried and convicted while they were subject to court-martial jurisdiction. See e.g., Carter v. McClaughry, 183 U.S. 365, 22 S.Ct. 181, 46 L.Ed. 236 (1902). Restricted to their facts, these cases do not control the question of whether a valid discharge prior to findings and sentence cuts off authority to try a former soldier by court-martial. We are aware of no ease where the principle of attached jurisdiction has been applied to justify the continuance of court-martial jurisdiction over a soldier who has been unconditionally discharged prior to trial. Nevertheless, the government urges us to extend the principle of continuing attached jurisdiction to justify the court-martial of persons over whom jurisdiction has attached, but who have been discharged prior to trial. We believe that Speller does not empower this court to judicially sanction the extension of military jurisdiction to a category of ex-soldiers who have not yet been convicted or sentenced by courts-martial. Accordingly, we decline to apply the principle of continuing attached jurisdiction in this case.
Our reluctance to extend Speller to pretrial discharge cases is motivated in part by our belief that Congress expects that a valid and unconditional discharge issued upon a soldier’s ETS will extinguish court-martial jurisdiction. An analysis of Article 3, UCMJ, supports this conclusion. In amending Article 3(a), UCMJ, following the Toth decision, Congress expressly created court-martial jurisdiction over a discharged person for pre-discharge offenses only if the dischargee subsequently regains some form of *595status subject to the UCMJ. See United States v. Gallagher, 7 U.S.C.M.A. 506, 22 C.M.R. 296, 1957 WL 4629 (1957). Moreover, Article 3(b), UCMJ, creates court-martial jurisdiction over a discharged person for pre-discharge offenses only if the dischargee is convicted by court-martial of obtaining his discharge by fraud. While Congress has explicitly extended military jurisdiction for predischarge offenses to the two categories of dischargees defined in Article 3(a) and (b), UCMJ, Congress has not relied on the notion of attached jurisdiction by expressly authorizing jurisdiction over soldiers, who, like the petitioner, have been lawfully discharged prior to findings but have not rejoined the military.
Accordingly, we are not prepared to conclude that Congress contemplated the broad and unprecedented judicial extension of court-martial jurisdiction that the government asks us to apply in this case. It is for Congress to adopt such a rule, particularly in light of the reality that any “action with a view to trial,” including the mere imposition of restraint, is sufficient to “attach” jurisdiction under military case law. R.C.M. 202(c). Thus, the class of unconvicted ex-soldiers over whom court-martial jurisdiction has “attached” while they were in a military status is large and amorphous. Under these circumstances, we believe that it is for Congress to explicitly decide which unconvicted discharged persons the military may pursue in the exercise of “the least possible power adequate” to regulate the Armed Forces. See Toth, 350 U.S. at 23, 76 S.Ct. at 8.
In reaching our decision not to judicially extend the application of Speller, we have carefully considered the discussion of continuing attached jurisdiction found in United States v. Hutchins, 4 M.J. 190 (C.M.A.1978), and its progeny. United States v. Douse, 12 M.J. 473 (C.M.A.1982); Smith, 4 M.J. at 267. Like Speller, the foregoing cases are not factually on point. They deal with soldiers in a jurisdictional status that is specifically provided for by statute, namely Article 2(a), UCMJ. Nevertheless, these cases contain dicta suggesting that in personam jurisdiction to try an accused, once attached, is not extinguished by a subsequent valid discharge from military service. Tracing the cited authority for this dicta to its antecedent, we are led to Colonel Winthrop’s explanation of the doctrine of continuing jurisdiction. William Winthrop, Military Law and Precedents 90-91 (2d ed., 1920 Reprint). However, we find that Winthrop’s discussion of continuing jurisdiction’s scope does not go as far as is submitted by those who quote him. He describes legal precedent establishing “continuing jurisdiction” only in terms of authorizing military officials to hold a soldier beyond his ETS, an assertion of jurisdiction over non-discharged soldiers that is expressly provided for by Congress in Article 2(a)(1), UCMJ. Moreover, Winthrop has more to say on discharges later in his treatise, when he states that the legal effect of a discharge is to separate a soldier “finally” from the service, and, whether he be a “deserter, an offender in arrest or on trial, or a convict,” he leaves the service in good standing. Winthrop at 550.
We hold, therefore, that a valid discharge of a soldier from the Army prior to trial operates as a formal waiver and abandonment of court-martial in personam jurisdiction, whether or not such jurisdiction had attached prior to discharge. Having decided that attached jurisdiction is severed by a valid discharge, it remains for us to decide whether the petitioner’s discharge was valid.
B. Incomplete Discharge
Title 10, United States Code, section 1168 (1988) provides that a member of the armed forces may not be discharged until his discharge certificate and his final pay are “ready for delivery to him.” Although a discharge certificate was actually delivered to petitioner, the government asserts that the petitioner’s discharge is invalid, because the Army was not prepared to deliver his final pay until it had been audited by the Defense Finance and Accounting Service (DFAS), a procedure usually completed thirty days after discharge. This contention is without merit. Granted, a valid discharge from active duty requires that a discharge certificate and final pay be ready for delivery. 10 U.S.C. § 1168 (1988). However, in the context of an ETS discharge, we construe this *596statute to mean that the military services must ensure that a soldier who has faithfully performed his service obligation is not discharged without a proper accounting of his final pay. The statute was not intended as a means of retaining court-martial jurisdiction when the government cannot or will not meet its obligation to timely deliver the soldier’s final pay. Moreover, we are satisfied from the record that EUSA discharging officials recognized their obligation under the statute. Their practice is to discharge soldiers after their final pay is computed and examined at the installation level. They do not hold soldiers beyond their ETS for the DFAS audit. Accordingly, we are satisfied that the petitioner’s discharge was issued in full compliance with 10 U.S.C. § 1168 (1988).
In reaching our conclusion, we have carefully considered United States v. King, 42 M.J. 79 (1995), which construed the requirements of the statute in the context of the processing necessary for an early discharge. Petitioner was not receiving an early discharge; he had reached his ETS. We do not believe King stands for the proposition that the Army must retain a soldier beyond his ETS in order to complete a redundant audit. Accordingly, we find that the requirement for a final audit routinely conducted thirty days after a soldier’s ETS was not a prerequisite to discharge under 10 U.S.C. § 1168 (1988). That requirement was no more than a supplemental procedure established solely for the convenience of the Army.
C. Fraud
The government asserts that the petitioner’s discharge should be disregarded because it was obtained by fraud. Though the military judge did not find actual fraud, he found facts that suggest that the petitioner may have obtained his discharge through a material misrepresentation. Congress has established the procedures of Article 3(b), UCMJ, as the exclusive means by which military authorities may exercise in person-am jurisdiction over pre-discharge offenses committed by those who have subsequently obtained facially valid discharges. Those procedures require a predicate conviction of the fraudulent discharge at a separate trial by court-martial before jurisdiction over predischarge offenses may be asserted. See United States v. Reid, 43 M.J. 906 (Army Ct.Crim.App.1996). We find no basis to conclude that Congress intended for military courts to permit the impeachment, on the basis of fraud, of a facially valid discharge through any other procedure than that provided for in Article 3(b), UCMJ. Compare United States v. Cole, 24 M.J. 18, 28 (C.M.A.) (Everett, C.J., dissenting),3 cert, denied, 484 U.S. 828, 108 S.Ct. 97, 98 L.Ed.2d 58 (1987).
D. Mistake of Fact
The government asserts that the discharge authority would not have issued the petitioner’s discharge had she been effectively placed on notice that court-martial charges were pending. From this, the government wishes us to conclude that the discharge is void because either the discharging official was operating under a mistake of fact, or the discharging official’s action was contrary to the actual (albeit undisclosed) intent of the convening authority and was-therefore not an “informed” action.
A commander desiring to hold a soldier beyond his ETS for purposes of court-martial should make that matter officially known by means of a flagging action. Army Reg. 600-8-2, Suspension of Favorable Personnel Actions (Flags)(1 Mar. 88)[herein-after AR 600-8-2], It may be inferred from the absence of a flagging action that no prosecutorial intent exists. Prior to issuing the discharge in this case, the discharging official determined that the petitioner’s records had not been flagged. She rationally concluded both that no court-martial charges were pending, and that she was obliged to issue the ETS discharge. Though her belief that no charges were pending was incorrect, under the circumstances, it was the only rational conclusion based on the entire record. Acting under the authority delegated to *597her, she discharged the petitioner in accordance with discharge procedures established by the Army. Such an attenuated mistake of fact “does not vitiate a purported discharge, even [if] the mistaken fact was necessary to the authority of the discharging officer to order the discharge.” Banner, 22 C.M.R. at 515.
The government also argues that because the discharge official mistakenly believed that the petitioner was not pending court-martial charges, the Army did not actually intend for the discharge to unconditionally sever all ties between the Army and the petitioner. The discharge official and the convening authority operated in different chains of command. Thus, we cannot logically impute the convening commander’s intent to the discharge official. Judging the Army’s intent through the eyes of the person authorized by the Army to discharge the petitioner, we are satisfied that at the time the discharge was issued, the Army fully intended the discharge to unconditionally sever its relationship with the petitioner, and vice versa.
E. Mistake of Law
The military judge ruled as a matter of law that the pendency of court-martial charges rendered the petitioner ineligible for discharge and deprived the discharging official of actual authority to issue the petitioner’s ETS discharge. In support of the military judge’s ruling, the government argues that the petitioner’s discharge is invalid because the discharge authority acted beyond the scope of the authority delegated to commanders by the Secretary of the Army under AR 635-200, Enlisted Personnel (17 Oct. 90)[hereinafter AR 635-200],
A soldier is normally entitled to receive an unconditional discharge upon his ETS, although this entitlement is not absolute. Commanders considering trial by court-martial after a soldier’s ETS have the authority to withhold an ETS discharge. Congress has explicitly extended court-martial jurisdiction in Article 2(a)(1), UCMJ, to include persons “awaiting discharge after expiration of their terms of enlistment.” Thus, commanders considering trial by court-martial after the expiration of a soldier’s term of service may withhold an ETS discharge until the charges are resolved. In such a case, a person subject to the code remains in that status until the person is unconditionally released from active duty through the formalities of a discharge. United States v. Poole, 30 M.J. 149 (C.M.A.1990); United States v. Hutchins, 4 M.J. 190,192 (C.M.A.1978).
A regulatory scheme is in place to withhold and prevent accidental ETS discharges in situations like this. The Secretary of the Army has promulgated two regulations establishing Army policy and procedure regarding the discharge of enlisted soldiers and their retention after their ETS: AR 635-200 and AR 600-8-2. Together, these regulations establish the general principles which guide the Army in its management of ETS discharges for persons awaiting trial by court-martial.
Normally, a soldier is eligible for discharge upon his ETS. AR 635-200, para. 4-2. However, Army policy stated in AR 635-200, para. l-24a, recognizes that a soldier who is otherwise eligible for ETS discharge “may” be lawfully retained after his term of service has expired when court-martial action is contemplated. Further, “[a] soldier who is awaiting trial ... by court-martial when he ... would otherwise be eligible for discharge ... will not be discharged ... until final disposition of the court-martial charges.” AR 635-200, para. l-24b. The Army flagging procedure is expressly designed to “guard against the accidental execution” of discharges for soldiers not in good standing. AR 600-8-2, paras. 1-8 and 1-14. Importantly, AR 600-8-2, para. 1-16, provides that soldiers will not be retained beyond their ETS unless “authority [is] obtained from the GCMCA [General Court-Martial Convening Authority] to extend the ETS.” Construing these related policy provisions together with applicable law, we conclude that the admonition against discharge contained in AR 635-200, para. l-24b, does not restrict a soldier’s eligibility for ETS discharge until the discharge official has received authority from the GCMCA to withhold the discharge.
*598Accordingly, we are satisfied as a matter of law and regulation that arraignment by court-martial does not operate automatically either to restrict a soldier’s eligibility for ETS discharge or to limit the actual authority of a properly appointed discharge official to issue a valid ETS discharge. “Separations will be accomplished by the [transition point by] ... processing the soldier for separation ... per the separation orders issued by the appropriate commander.” AR 635-200, para. 4-5. To limit the authority of the discharge official to issue a valid discharge, the GCMCA must do two things: (1) take appropriate action with a view towards court-martial; and, (2) take steps to effectively authorize the discharge authority to withhold the ETS discharge. The method specified in Army regulations for communicating the lawful authority to withhold an ETS discharge is a flagging action with the imprimatur of the GCMCA.
Inasmuch as the GCMCA had not effectively provided the discharging official with authority to extend the petitioner’s ETS, we hold that the petitioner remained eligible for discharge, and the discharge authority had actual authority under Army regulations to unconditionally discharge the petitioner. Under the circumstances of this case, we find no “mistake of law” that would void petitioner’s discharge. See United States v. Banner, 22 C.M.R. 510, 1956 WL 4820 (A.B.R.1956); United States v. Garvin, 26 M. J. 194 (C.M.A. 1988) (discharge void, barred by statute).
TV. DECISION
This case presents unprecedented circumstances combining the uncoordinated actions of a convening authority and a discharging official operating within different chains of command, each independently authorized by law to carry out their respective functions to completion. Accordingly, our decision should be strictly limited to the facts of this case.
We find that the petitioner was validly discharged by competent authority under honorable conditions at his ETS and prior to trial. The valid discharge operated to unconditionally sever all ties between the military and the petitioner. Accordingly, the discharge operated to sever the jurisdiction of the pending court-martial at the time of discharge. Except for the provisions of Article 3(b), UCMJ, Congress has not statutorily extended court-martial jurisdiction to persons situated like the petitioner. Therefore, we hold that the petitioner is a civil person not subject to court-martial jurisdiction for offenses committed prior to his discharge, unless he is later convicted by court-martial of obtaining his discharge by fraud in violation of Article 83, UCMJ, or subsequently reenlists in the Army.
The charges are dismissed, without prejudice, and may be re-preferred should jurisdiction over these offenses later be established under the provisions of Article 3(a) or (b), UCMJ.
Judge TOOMEY concurs.. The petitioner filed the original petition, a brief in support of the petition, and a reply to the *593answer to show cause.
. The military judge’s complete findings of fact and conclusions of law are appended to this opinion.
. Chief Judge Everett opines that "Once military criminal jurisdiction attaches, it is not lost by a later change in the accused’s status." Cole, 24 M.J. at 28 (Everett, C.J., dissenting). We very respectfully do not believe that this conclusion is supported by the plain meaning of Article 3(b), UCMJ.