I respectfully dissent. I agree with the majority that the issue of the admissibility of Agent Prodan’s testimony was preserved for appellate review, and that the trial judge erred in admitting Prodan’s testimony without determining that the evidence was reliable. I disagree, however, that the admission of that testimony was harmless error.
As I read the majority opinion, it does not view Prodan’s testimony in light of the properly admitted evidence in the record to determine whether this erroneously admitted evidence could have contributed to the jury’s verdict. Chapman v. California, 386 U.S. 18, 87 S.Ct. 824, 17 L.Ed.2d 705 (1967). *392Instead, the majority posits that the jury made three critical factual determinations:
(1) that the DNA evidence established respondent was the source of the protein found in the victim’s vaginal vault;
(2) that the DNA was not present as a result of consensual sexual activity between the victim and respondent; and
(3) that the inmates’ testimony of respondent’s purported confession was credible.
The majority then determines that Prodan’s testimony could not have contributed to the verdict obtained. I disagree.
The majority states Prodan’s testimony did not “promote the validity of the DNA results or boost the credibility of the inmates’ testimony about Respondent’s confession.” Admittedly Prodan provides no direct evidence regarding the DNA evidence, but as explained below his erroneously admitted “expert” testimony bolstered both the probability that respondent was the source of the DNA as well as the inmates’ testimony relating respondent’s “confession.” In my view, the critical harmless error question is whether Prodan’s erroneously admitted testimony bolstered the State’s contention that respondent was the victim’s rapist and killer.
Prodan’s testimony that the victim probably knew her attacker bolstered the inmates’ testimony that respondent had confessed to an ongoing relationship with the victim. Further, Prodan’s theory that the victim and her assailant were acquainted served to identify respondent as a likely perpetrator in light of his statement to police that he had been in the apartment, as well as the testimony of the victim’s roommate and another witness that the victim was familiar with respondent. Accordingly, Prodan’s testimony cannot be said to be harmless, especially given the relative statistical weakness of the DNA evidence.
In respondent’s “confession,” he did not acknowledge any sexual contact on the night of the killing, but rather told the inmates that he went to the victim’s apartment in a jealous frame of mind, having learned his “girlfriend” was sexually involved with a mutual friend. According to the inmates’ testimony, respondent confessed his jealous frame of mind escalated to rage as he questioned the woman about the relationship, and resulted in the victim’s stabbing. Prodan’s *393testimony that “the assailant most likely did not come to the apartment for the purpose of killing the victim” bolsters this part of respondent’s confession. Prodan’s theory that the perpetrator intended to sexually humiliate the victim by posing her body is consistent with respondent’s confession that he went to the apartment after learning of the victim’s sexual infidelity, and that her responses to his questions about this infidelity led to her assault. Prodan’s improperly admitted evidence goes directly to the only issue the jury had to decide: not whether crimes had occurred, but was respondent the perpetrator.
Unlike the majority, I believe that Prodan’s improperly admitted “expert” testimony cannot be deemed harmless error. Improper “expert” evidence which goes to the heart of the case is not harmless. See State v. Ellis, 345 S.C. 175, 547 S.E.2d 490 (2001); see also State v. Douglas, 380 S.C. 499, 671 S.E.2d 606 (2009) (Pleicones, J., dissenting). I would therefore reverse and remand the case for a new trial.