(dissenting).
I respectfully dissent. In my view, the error in admitting stale allegations of child sexual abuse into Ness’s trial is reversible error. I also write separately to address the relevance of prior bad acts that are remote in time from the charged offense, and the court of appeals’ reading of our decision in State v. Wermerskirchen, 497 N.W.2d 235 (Minn.1993).
I.
The district court did not make a specific finding as to whether the Caron evidence was relevant and the court of appeals, relying on Wermerskirchen, ■ concluded that it was.
In Wermerskirchen, the defendant was charged with sexually abusing his nine-year-old daughter four years before the trial. Id. at 238. At trial, Spreigl evidence of the defendant’s sexual abuse of his stepdaughter, J.L., and two of his nieces, C.S. and J.R., was admitted. Some of the incidents involving the sexual abuse of J.L. occurred as many as nine years before the defendant’s trial. The incidents involving J.R. occurred over a period of time between 9 and 17 years before the trial. And the incidents of abuse involving C.S. occurred as many as 19 years before the trial. Id. at 237. Significantly, in footnote 3 of Wermerskirchen we said:
Often the passage of time, while superficially significant, turns out to be without real significance * * * if it turns out that the defendant was in prison in the interval between the prior offense and the current offense and was incapacitated from committing crime [or] if the older offense is part of a “pattern” of *692similar misconduct occurring over a number of years.
Id. at 242 n. 3. We held that the Spreigl evidence relating to J.L., J.R., and C.S. was properly admitted at trial because it was part of a pattern of sexual abuse by the defendant. Id. at 242-43.
Beyond Wermerskirchen, we have, in a number of other cases, upheld the admission of evidence of other crimes, wrongs, or acts that occurred years before the charged crime.1 Those cases are consistent with what we said in Wermerskirchen in that the evidence admitted was part of a pattern of conduct by the defendant and/or the defendant was in jail or otherwise incapacitated from engaging in similar conduct during the intervening years.2 Recently, in State v. Washington, we held that the district court did not abuse its discretion when it admitted evidence of sexual abuse that occurred 16 years before the last sexual abuse of the complainant. State v. Washington, 693 N.W.2d 195, 200 (Minn. 2005). We explained in Washington that our cases, including Wermerskirchen, show that:
[A] district court, when confronted with an arguably stale Spreigl incident, should employ a balancing process as to time, place, and modus operandi: the more distant the Spreigl act is in terms of time, the greater the similarities as to place and modus operandi must be to retain relevance. They also show that concerns about acts that are remote in time are lessened where the defendant spent a significant part of that time incarcerated and other intervening acts fend to bolster the prior act’s relevance and materiality. Such concerns are also lessened if the defendant was actually convicted of a crime based on the prior bad acts because the process of securing a conviction (i.e., obtaining a defendant’s plea of guilty or gathering and presenting evidence at a trial) reduces the actual prejudice to the defendant of being required to defend against claims concerning acts that occurred years ago.
Id. at 202. We also noted that, “[i]n general, the prior acts become less relevant as time passes. Thus, the greater the time gap, the more similar the acts must be to lessen the likelihood that the Spreigl evidence will be used for an improper purpose.” Id. at 201 (internal quotations and citations omitted). In Washington, we affirmed the district court’s admission of the Spreigl evidence because, while there were no acts of sexual misconduct during the intervening years, the defendant had been convicted of the Spreigl incident, had been incarcerated for over half of that time period, and the Spreigl incidents were “strikingly similar in modus operandi to the charged acts.” Id. at 202-03.
In relying on our decision in Wermer-skirchen, the court of appeals characterized it as a case in which this court admitted 19-year-old Spreigl evidence “to establish a pattern of improper sexual conduct.” State v. Ness, No. A03-1187, 2004 WL 1444952 at *2 (Minn.App. June 29, 2004). That characterization is correct as far as it goes. Unfortunately, the court of appeals, in. analyzing the present *693case, did not go beyond that characterization. In Wermerskirchen we explained that under two circumstances the significance of the passage of time is diminished: (1) when the defendant was in prison or otherwise incapacitated for all or a portion of the interval between the occurrence of the Spreigl incident and the charged offense; and (2) when “the older offense is part of a pattern of similar misconduct occurring over a number of years.” 497 N.W.2d at 242 n. .3 (citations omitted). Interestingly, the court of appeals’ decision did not quote that portion of footnote 3 of Wermerskirchen that discussed the significance of the Spreigl incident being a part of a pattern or the effect of incarceration or other incapacity on the passage of time.3 It is interesting *694because in its decision the court of appeals did not analyze whether the Caron evidence was part of a pattern of sexual abuse or whether there was some incapacity that made the passage of time lose its significance. Indeed, a fair reading of the court of appeals’ decision in Ness suggests that the court read Wermerskirchen as creating a bright-line rule that in sexual abuse cases Spreigl evidence will always have real significance, no matter how much time passed between the other occurrence and the charged offense. We did not, however, create a bright-line rule. Nothing in Wermerskirchen indicates that the passage of time does not have real significance when the gap in time between the charged offense and the Spreigl incident is great and there is no pattern of sexual misconduct during the intervening years, and the defendant was not in prison or otherwise incapacitated from engaging in sexual misconduct. Thus, courts must examine the facts of each case to determine the significance, if any, of the passage of time.
The facts of this case are distinguishable from Wermerskirchen, Washington, and the other cases in which we have upheld the admission of Spreigl evidence in the face of a claim that the passage of time made that evidence lose its relevance. None of those cases involved other crimes, wrongs, or acts that occurred three decades before the charged offense. More importantly, here, during the 35 years in question, Ness was an elementary school teacher, a principal, and a volunteer art instructor who had constant contact with children. Yet the record does not contain any evidence indicating that during that time Ness either engaged in or was alleged to have engaged in any sexual misconduct with children. Thus, the record does not indicate any pattern of sexual misconduct during that time period despite Ness’s continued exposure to children. Nor is there any evidence that Ness was incapacitated from engaging in sexual misconduct during that period of time. If anything, the record suggests that Ness was in a position to engage in such misconduct but did not. Thus, in this case, I conclude that the passage of time has real significance in that the incidents admitted as Spreigl evidence were stale and too old to be relevant.
II.
A defendant is entitled to a new trial when the district court erroneously admits Spreigl evidence if “there is a reasonable possibility that the wrongfully admitted evidence significantly affected the verdict.” State v. Bolte, 530 N.W.2d 191, 198 (Minn.1995); see also State v. Post, 512 N.W.2d 99, 102 n. 2 (Minn.1994). In cases tried to a judge, I presume that the judge knows the law and restricts evidence of limited admissibility to its proper scope and purpose. See Minn. R. Evid. 105. Therefore, I presume that in admitting the Spreigl evidence involving Caron, the district court did not use it for an improper purpose. Even making this presumption, it must still be determined whether there is a reasonable possibility that the use of the improperly admitted evidence “significantly” affected the verdict. According to the district court, the main issue in the case was *695whether Ness touched E.M. with sexual or aggressive intent. In the court’s memorandum of law it found that the testimony-offered by E.M. about the incident was “very credible.” The court also found that Voigt’s testimony about the incident was consistent with E.M.’s testimony, added credibility to E.M.’s testimony, and was “credible in and of itself.”
Based on that testimony, the district court found that “[NessJ’s actions * * *, along with the nature and number of times he touched E.M.’s intimate parts, is by itself sufficient to show beyond a reasonable doubt that [Ness] acted with sexual or aggressive intent.” Notwithstanding that finding, the district court specifically indicated that the Spreigl evidence involving Caron “bolstered” its finding that Ness touched E.M. with sexual or aggressive intent. Given this statement by the district court, I conclude that there is a reasonable possibility that the wrongfully admitted Caron evidence significantly affected the verdict. Had the evidence not had a significant effect on the verdict, there would have been no reason, given the district court’s other findings, to have noted that the Caron evidence bolstered its findings on intent.
Even if there arguably was not a reasonable possibility that the verdict was affected by the Caron evidence, I believe we should nevertheless reverse in the exercise of our supervisory powers. We should do so to make clear that district courts may not admit irrelevant, inherently prejudicial evidence of this nature.
I would therefore reverse the court of appeals and remand for a new trial.
. See State v. Crocker, 409 N.W.2d 840 (Minn.1987) (nine-year-old offense was relevant and material because defendant was in prison during the nine-year interval); State v. Filippi, 335 N.W.2d 739 (Minn.1983) (Spreigl evidence of defendant's participation in a similar crime four years earlier was relevant and material to the charged offense because defendant was in prison for most of the four-year period); State v. Bolts, 288 N.W.2d 718 (Minn.1980) (Spreigl evidence was admissible when defendant was in prison during the four-year interval between the charged offense and prior crime evidence).
. See infra note 3.
. That footnote in its entirety reads:
Defendant complains about the lack of closeness in time between the charged offense and the other crimes. We have never held that there must be a close temporal relationship between the charged offense and the other cñme. In State v. Filippi, 335 N.W.2d 739, 743 (Minn.1983), we said, "In determining relevancy, we have generally required that the other crime be similar in some way — either in time, location, or mo-dus operandi — to the charged offense, although 'this’ [sic] of course, is not an absolute necessity. ” Often the passage of time, while superficially significant, turns out to be without real significance. Thus, e.g., the passage of a number of years may be without real significance if it turns out that the defendant was in prison in the interval between the prior offense and the current offense and was incapacitated from committing crime. See State v. Filippi, 335 N.W.2d 739, 743-44 (Minn.1983), and cases cited therein. Also, the passage of a number of years may be without real significance if the older offense is part of a "pattern" of similar misconduct occurring over a number of years. State v. Anderson, 275 N.W.2d 554 (Minn.1978). An illustrative case is State v. Crocker, 409 N.W.2d 840 (Minn.1987), where, in the prosecution of the defendant for raping a college student, we upheld the admission of evidence that 9 years earlier, in 1977, the defendant had sexually assaulted a 7-year-old girl. We said:
The fact that defendant committed a sex offense in 1977, was in prison for most of the next 9 years, and then in 1986 committed two sex offenses before the current offense shows a relevant pattern of sexually assaultive conduct. The fact that the 1977 offense involved the sexual assault of a 7-year-old girl rather than a woman or a sexually mature young woman should not necessarily make a difference. * * * The 1977 offense in question involved the opportunistic sexual assault of a vulnerable 7-year-old girl during a brief period of time when defendant was alone with her. The January 1986 offense involved the opportunistic attempt to assault a vulnerable 15-year-old stepdaughter sexually during a brief period of time when the girl's mother went to the grocery store on a quick errand. The assault on [a friend of the complainant shortly before the defendant raped the complainant] was a similar opportunistic assault on a young woman who was in a temporarily vulnerable position. The charged offense occurred under similar circumstances.
409 N.W.2d at 843. See also State v. Rainer, 411 N.W.2d 490 (Minn.1987), where we upheld the admission of evidence of 16 to 19-year-old incidents which showed a repeating pattern of very similar conduct. The ultimate issue is not the temporal relationship but relevance. "Older” offenses sometimes are relevant, sometimes not. Relevance generally must be determined by the trial court, with review limited to whether the trial court abused its discretion.
The other crimes here occurred between 1970 and 1982, whereas the current offense occurred in 1989. The trial court concluded that this time lapse did not require exclusion because, as the trial court put it, "The other Spreigl factors allow for the admission of this evidence.” Specifically, the trial court said:
In each case the incidents have a similar modus operandi on the part of the Defendant, i.e., rubbing buttocks and touching breasts. The incidents also have a similarity in location, i.e., in or at the home of the alleged victim. The incidents involved * * * inappropriate sexual contact with young female relatives of the ages of 10 to 13 years of age, and the victims *694were all pre-adolescents at some point in the alleged encounters.
The victim in this case was also a pre-adolescent being approximately 8 years and 3 months old at the time of the alleged incidents in April, 1989. The type of conduct, the location of the conduct and the youth of the victims is similar to the crime charged. Further, the sexual intent of the incidents is clear.
Wermerskirchen, 497 N.W.2d at 242 n. 3 (emphasis added). The italicized portion of the footnote was quoted in the court of appeals’ Ness opinion.