Opinion by Judge GOODWIN; Dissent by Judge TALLMAN.
GOODWIN, Circuit Judge:Petitioner Gilbert C. Brown alleges that his constitutional rights were violated when a statute, enacted after the commission of his crimes, was applied by the Oregon State Board of Parole and Post-Prison Supervision (“Board”) to postpone his parole release date. He contends that this violation of the Ex Post Facto Clause constitutes a constitutional injury that compels reversal of the district court’s denial of his petition for habeas corpus. We agree.
BACKGROUND
I. Oregon’s Statutory Scheme for Postponing Parole Release
Brown challenges the application of an Oregon statute, passed after he committed his crimes, which changed the method by which the Board decided whether to postpone a prisoner’s parole date. Before December 1993, Oregon Revised Statutes § 144.125 provided:
If a psychiatric or psychological diagnosis of present severe emotional disturbance such as to constitute a danger to the health or safety of the community has been made with respect to the prisoner, the Board may order the postponement of the scheduled parole release until a specified future date.
Or.Rev.Stat. § 144.125(3) (1991).
In 1993, the Oregon legislature amended § 144.125(3) to read:
If the Board finds the prisoner has a mental or emotional disturbance, deficiency, condition or disorder predisposing the prisoner to the commission of a crime to a degree rendering the prisoner a danger to the health or safety of the community, the Board may order the postponement of the scheduled parole release until a specified date.
Or.Rev.Stat. § 144.125(3)(a) (1993).
II. Factual Background
Petitioner Brown was convicted in 1982 of four counts of sodomy and one count of rape for assaulting his daughter. Brown received indeterminate sentences of sixty years for these crimes.
In a July 1995 psychological evaluation, Dr. Robert Davis found that Brown “does exhibit some signs of emotional disturbance, particularly of a passive-dependent nature,” but did not diagnose Brown with any disease or disorder. Dr. Davis concluded his evaluation by stating: “I do not find that he represents a severe or extreme emotional disturbance or that he is a danger to the health and safety of others in the community at the present time.” In a September 1995 action, the Board postponed Brown’s parole release date for twenty-four months, finding that he was a “danger to the health or safety of the community.” In light of an Oregon appel*1092late court opinion interpreting the pre-1993 version of § 144.125, the Board revised its findings in a February 1997 administrative review to conclude that Brown suffered from a “present severe emotional disturbance.”
Brown filed a state habeas petition in 1996, contending that the Board’s 1995 action violated the Ex Post Facto Clause by applying the post-1993 version of § 144.125(3). An Oregon circuit court dismissed Brown’s habeas claim for the reasons given in the state’s Motion to Dismiss as well as those in Adams v. Thompson, Marion County Circuit Ct. Case No. 96C11605, and his appeals were dismissed as moot. Brown filed a federal habeas petition in April 1999, alleging constitutional violations under the Ex Post Facto Clause and the Due Process Clause of the Fourteenth Amendment. While his petition was pending in the district court, the Oregon Supreme Court decided a case that made the mootness determination void, and the district court sent the case back to the state courts for further review. The Oregon appellate and Supreme Court again denied Brown’s state habeas petition. The district court reactivated Brown’s federal habeas case in October 2002 and also denied his petition. Brown appeals that denial, alleging only an Ex Post Facto violation.
STANDARDS OF REVIEW
A district court’s denial of a petition for habeas corpus is reviewed de novo. Himes v. Thompson, 336 F.3d 848, 852 (9th Cir.2003). Because Brown filed his federal habeas petition after the effective date of the Antiterrorism and Effective Death Penalty Act (“AEDPA”), Pub. L. No. 104-132, 110 Stat. 1214 (Apr. 24,1996), its provisions apply to his case. Fuller v. Roe, 182 F.3d 699, 702 (9th Cir.1999) (per curiam).
Under AEDPA, a habeas petitioner may be granted relief if the state court’s decision “resulted in a decision that was contrary to, or involved an unreasonable application of, clearly established Federal law, as determined by the Supreme Court of the United States.” 28 U.S.C. § 2254(d)(1). To show that the state court unreasonably applied federal law, a petitioner must demonstrate “that the state court’s application of Supreme Court precedent to the facts of his case was not only incorrect but ‘objectively unreasonable.’ ” Davis v. Woodford, 333 F.3d 982, 990 (9th Cir.2003) (quoting Woodford v. Visciotti, 537 U.S. 19, 25, 123 S.Ct. 357, 154 L.Ed.2d 279 (2002)); see also Williams v. Taylor, 529 U.S. 362, 409, 120 S.Ct. 1495, 146 L.Ed.2d 389 (2000) (O’Connor, J., concurring).
The last reasoned state court decision in Brown’s case was the Oregon Court of Appeals’ decision on a petition for reconsideration, in which it adhered to its original dismissal of Brown’s state habeas petition. Brown v. Thompson, 177 Or. App. 530, 34 P.3d 735 (2001); see Robinson v. Ignacio, 360 F.3d 1044, 1055 (9th Cir.2004) (“[T]he federal court should review the ‘last reasoned opinion’ by a state court ....”) (citing Avila v. Galaza, 297 F.3d 911, 918 (9th Cir.2002)). In rejecting Brown’s Ex Post Facto arguments, the Brown court stated simply that “the record d[oes] not support petitioner’s contention.” 177 Or.App. at 533, 34 P.3d 735. The opinion to which the court adhered was an affirmance without opinion. Brown v. Thompson, 176 Or.App. 169, 31 P.3d 453 (2001). Because the Oregon courts have provided no ratio decidendi to review, or to which we can give deference, we employ the “objectively reasonable” test. In this situation, federal habeas courts accord the state court decisions less deference than in standard habeas cases. *1093Delgado v. Lewis, 223 F.3d 976, 982 (9th Cir.2000) (citing Delgado v. Lewis, 181 F.3d 1087, 1092-93 (9th Cir.1999)).
ANALYSIS
I. Controlling Supreme Court Law on Ex Post Facto Violations
The Constitution’s aversion to retroactive legislation finds expression in several constitutional provisions. See Lynce v. Mathis, 519 U.S. 433, 439 n. 12, 117 S.Ct. 891, 137 L.Ed.2d 63 (1997). Separate clauses prohibit the passage of Ex Post Facto laws by Congress, U.S. Const. art. I, § 9, cl. 3, and by the states, id. art. I, § 10, cl. 1. The Ex Post Facto clauses forbid the passage and application of laws that “ ‘retroactively alter the definition of crimes or increase the punishment for criminal acts.’ ” Calif. Dep’t of Corr. v. Morales, 514 U.S. 499, 504, 115 S.Ct. 1597, 131 L.Ed.2d 588 (1995) (quoting Collins v. Youngblood, 497 U.S. 37, 41, 110 S.Ct. 2715, 111 L.Ed.2d 30 (1990)).
The Supreme Court has established a two-part test for assessing Ex Post Facto claims. A law violates the Ex Post Facto Clause if it is 1) retroactive — it “applies] to events occurring before its enactment,” Weaver v. Graham, 450 U.S. 24, 29, 101 S.Ct. 960, 67 L.Ed.2d 17 (1981); and 2) detrimental — it “produces a sufficient risk of increasing the measure of punishment attached to the covered crimes.” Morales, 514 U.S. at 509, 115 S.Ct. 1597; Himes, 336 F.3d at 854. “The inquiry looks to the challenged provision, and not to any special circumstances that may mitigate its effect on the particular individual.” Weaver, 450 U.S. at 33, 101 S.Ct. 960; Nulph v. Faatz, 27 F.3d 451, 455-56 (9th Cir.1994). The Court has also specified that the statutory procedures should be compared “in toto to determine if the new may be fairly characterized as more onerous.” Dobbert v. Fla., 432 U.S. 282, 294, 97 S.Ct. 2290, 53 L.Ed.2d 344 (1977).
Not every law that “work[s] to the disadvantage of a defendant” violates the Ex Post Facto Clause. Dobbert, 432 U.S. at 293, 97 S.Ct. 2290. Changes that are merely procedural will withstand scrutiny, as will statutes that leave unaffected “ ‘[t]he crime for which the [ ] defendant was indicted, the punishment prescribed therefor, and the quantity or the degree of proof necessary to establish his guilt....”’ Id. at 294, 97 S.Ct. 2290 (quoting Hopt v. Utah, 110 U.S. 574, 589, 4 S.Ct. 202, 28 L.Ed. 262 (1884)). A law does not violate the clause if it “creates only the most speculative and attenuated risk of increasing the measure of punishment attached to the covered crimes.” Morales, 514 U.S. at 513, 115 S.Ct. 1597. A change in law that changes the way a parole board exercises its discretion does not necessarily offend the Ex Post Facto Clause. Garner, 529 U.S. at 253, 120 S.Ct. 1362. The Court has been careful to explain that there is no single formula for identifying Ex Post Facto violations. Id. at 252, 120 S.Ct. 1362; Morales, 514 U.S. at 509, 115 S.Ct. 1597. The inquiry “must be a matter of degree.” Morales, 514 U.S. at 509, 115 S.Ct. 1597 (quoting Beazell v. Ohio, 269 U.S. 167, 171, 46 S.Ct. 68, 70 L.Ed. 216 (1925)) (emphasis in original).
In engaging in Ex Post Facto analysis, a federal court “ ‘accepts the meaning ascribed to [state statutes] by the highest court of the state.’ ” Souch v. Schaivo, 289 F.3d 616, 621 (9th Cir.2002) (quoting Murtishaw v. Woodford, 255 F.3d 926, 965 (9th Cir.2001)); Himes, 336 F.3d at 852. This is also true in the context of habeas review. See Estelle v. McGuire, 502 U.S. 62, 67-68, 112 S.Ct. 475, 116 L.Ed.2d 385 (1991) (“[W]e reemphasize that it is not the province of a federal habeas court to reexamine state-court de*1094terminations on state-law questions.”); Mullaney v. Wilbur, 421 U.S. 684, 691, 95 S.Ct. 1881, 44 L.Ed.2d 508 (1975) (stating that “state courts are the ultimate expositors of state law”); Mendez v. Small, 298 F.3d 1154, 1158 (9th Cir.2002) (“A state court has the last word on the interpretation of state law.”) (citing McSherry v. Block, 880 F.2d 1049, 1052 (9th Cir.1989)).
11. Ex Post Facto Analysis
A. Retroactivity
Brown argues that because the Board applied to him the parole statute as amended in 1993, the Board violated his rights. The pre-1993 law required that postponement be based upon a psychiatric or psychological diagnosis of a present severe emotional disturbance. The 1993 amendment eliminated the requirement of a diagnosis and eliminated “present severe” from the definition of the qualifying mental disturbance. Or.Rev.Stat. § 144.125 (1993).
In Weidner v. Armenakis, 154 Or.App. 12, 959 P.2d 623 (1998) (en banc), withdrawn July 13, 1998, reasoning readopted and aff'd Merrill v. Johnson, 155 Or.App. 295, 964 P.2d 284 (1998),1 an Oregon appellate court determined that the pre-1993 version of § 144.125(3) assigned the Board the task of determining whether a prisoner suffered from a “present severe emotional disturbance such as to constitute a danger to the health and safety of the community.” Weidner, 154 Or.App. at 19, 959 P.2d 623; see Or.Rev.Stat. § 144.125(3) (1991). The statutory standard is a “legal one, rather than a medical one.” Weidner, 154 Or.App. at 16, 959 P.2d 623. However, “a psychiatric or psychological diagnosis is a prerequisite to the Board’s consideration ....” Id. at 19, 959 P.2d 623 (emphasis added). In a later opinion, the Oregon appellate court clarified that such a diagnosis “must provide the foundation for the Board’s finding that the emotional disturbance in question is ‘present’ and ‘severe.’ ” Christenson v. Thompson, 176 Or.App. 54, 61, 31 P.3d 449 (2001). After the 1993 amendments, no such diagnosis was necessary. Under the new version of the statute, the Board had wide discretion to postpone a prisoner’s parole release date. While it could take into account a professional diagnosis, it was not in any way constrained by it.
Brown’s psychological evaluation does not provide a foundation for the Board’s postponement of Brown’s parole release date. Nothing about Dr. Davis’s report supports the Board’s finding that Brown suffered from a present severe emotional disturbance indicating that he was a danger to the community. In fact, Dr. Davis’s evaluation plainly refutes such a conclusion. Nor does Dr. Davis provide a diagnosis of any disorder — a prerequisite, under the Oregon courts’ interpretation of pre-1993 § 144.125(3), to the Board’s finding that a prisoner suffers from a “severe” emotional disturbance.2 Cf. Christenson, 176 Or.App. at 60-61, 31 P.3d 449. In postponing Brown’s parole release date, the Board retroactively applied the post-1993 version of the statute, employing a discretionary scope it did not possess un*1095der the statute in effect when Brown committed his crimes.
B. “Significant risk”
The second prong of Ex Post Facto analysis asks whether the retroactive statute works to some significant disadvantage to a petitioner, creating a “significant risk” that the statute’s application will increase the length of incarceration. Garner, 529 U.S. at 255, 120 S.Ct. 1362; Morales, 514 U.S. at 510, 115 S.Ct. 1597. The respondent does not attempt to show that the alleged changes are not detrimental to Brown; instead, she denies that the difference between the pre and post-amendment versions of § 144.125(3) is a material difference.
The Supreme Court teaches us to examine the retroactive statute both on its face and in real-world practice. Garner, 529 U.S. at 255, 120 S.Ct. 1362 (“When the rule does not by its own terms show a significant risk, the respondent must demonstrate, by evidence drawn from the rule’s practical implementation by the agency charged with exercising discretion, that its retroactive application will result in a longer period of incarceration than under the earlier rule.”). We have previously stated that in absence of a disadvantage that affects prisoners in general, “an individual will satisfy the detriment requirement if he shows that it can ‘be said with assurance’ that he would have received less severe punishment under the prior scheme.” Nulph, 27 F.3d at 456 (quoting Dobberi, 432 U.S. at 294, 97 S.Ct. 2290).
When compared in toto with the pre-1993 version of § 144.125(3), the post-1993 statute creates a significant risk that prisoners will face longer periods of inearceration. The Oregon courts have twice held that retroactive application of posN1993 § 144.125(3) violates the Ex Post Facto Clause. In Meadows v. Schiedler, 143 Or.App. 213, 924 P.2d 314 (1996), an Oregon court of appeals held that the deletion of the word “severe” from § 144.125(3) made the standards for postponing a prisoner’s parole release date less strict. The court held that this violated the Ex Post Facto Clause, and remanded the case to the Parole Board to reconsider the habeas petitioner’s case in light of pre-1993 version’s requirements. 143 Or.App. at 219-20, 924 P.2d 314. A later decision held that Weid-ner’s interpretation could not be applied to prisoners who had committed their offenses during the effective period of an administrative rule restricting the Board’s authority to extend a parole release date. Peek, 160 Or.App. at 266, 980 P.2d 178.
The standard applied in Brown’s Board hearing enlarged the Board’s discretion beyond that granted in the pre-1993 statute, as interpreted by the Oregon courts.3 Under the old statute, a medical diagnosis was prerequisite to the Board’s postponing a prisoner’s parole release date. Weidner, 154 Or.App. at 19, 959 P.2d 623. The new standard gives the Board the discretion to find that a prisoner is a danger to the community despite the contrary conclusions of a psychological/psychiatric evaluation. This creates a “sufficient risk” of increased punishment to constitute violation of the Ex Post Facto Clause. Himes, 336 F.3d at 854. Had the Board applied the pre-1993 version of the statute, it could not have postponed Brown’s parole release date. It can “ ‘be said with assurance’ ” that Brown would have had a shorter period of incarceration. *1096Nulph, 27 F.3d at 456 (quoting Dobbert, 432 U.S. at 294, 97 S.Ct. 2290).
CONCLUSION
The Oregon courts’ dismissal of Brown’s Ex Post Facto claim was objectively unreasonable in light of the Supreme Court’s expressed guidance on Ex Post Facto cases. In postponing Brown’s parole release date, the Board retroactively applied a version of a parole statute enacted after Brown’s crimes to Brown’s detriment. Accordingly, the judgment denying Brown’s petition is
REVERSED.
. As noted in the full cite, Weidner was withdrawn soon after its publication because it became moot. It was readopted in full in Merrill, which is the precedential case, but the Oregon courts refer to Weidner. Peek v. Thompson, 160 Or.App. 260, 262 n. 2, 980 P.2d 178 (1999) (en banc). We follow that convention.
. The opinion from the Oregon Court of Appeal cited in the dissent is an affirmance without opinion, and accordingly, we are unable to discern what significance, if any, the Oregon court gave to Dr. Davis’s psychiatric evaluation.
. We do not reach the question whether the state courts' interpretation of § 144.125 (1991) constitutes "obvious subterfuge.” The denial of Brown's habeas petition must be reversed under any reading of the statute, including the interpretation given it by the Oregon courts.