Discussions concerning homosexual men currently focus on topics such as public health (e.g., AIDS and its transmission and prevention), identification processes, oppression and stigma, and the high prevalence of sexual abuse (Li Fangying et al., 2014; Liu Pei-Ling et al., 2016; Chuang & Lacombe-Duncan, 2016; Hall, 2018; Smit et al., 2012; Wohl et al., 2013; Wolitski & Fenton, 2011). In the past, homosexuality was unacceptable in mainstream society, and same-sex relationships underwent discrimination (Elipe et al., 2018). Perhaps because of this social context and the fact that sexual assault signifies conquest and power (Brownmiller, 1975), which is inconsistent with the image of homosexual men, research on the perpetration of sexual assault by homosexuals is limited. Homosexual men’s perpetration of sexual assault is not the same as male sexual assault on men. The latter perpetrators are not necessarily homosexuals, whereas a few of the former perpetrators may be able to assault women. Some studies have explored male sexual assault on men (Abdullah-Khan, 2008), such as male-on-male sexual assault in prisons; however, research is yet to be conducted on sexual assault perpetrated by homosexual men.Hirschi (1969) indicated that committing crimes is part of human nature. Under this point of view, it is understandable that gay men commit sexual assault. Studies have failed to investigate this issue; however, we believed this issue is worthy of research attention. Because the majority of sexual assaults are currently still performed by men (Police Administration of the Ministry of the Interior, 2017), this study explored sexual assaults perpetrated by homosexual men.Scholars such as Marshall and Barbaree (1990), Hall and Hirschman (1991), Ward and Beech (2006), and Seto (2019) have presented the integrated theory of sexual assault, which means that regardless of the subject (male or female) or object (child, adult, male, or female), sexual assault is within the explanatory scope of their theory. In other words, irrespective of the sex offender’s sexual orientation, causes behind the offence can be explained by high-level constructs. Accordingly, this study proposed the following three risk factors that partly explain reasons for sexual assault by homosexual men: experiences of sexual abuse, acceptance of rape myths, and gender biases. Experiences of sexual abuse constitute the social learning factor (undergoing sexual assault and then following the example), whereas rape myths and gender biases are cognitive factors. All the three factors are emphasized by the integrated theory of sexual assault. These factors are more instructive and cognitive in nature and therefore more changeable (compared with factors that are biological in nature). Hence, we chose them as the subject of our investigation to facilitate subsequent suggestions on sexual assault prevention.We assumed the following: (a) The experiences of sexual abuse, degree of rape myth acceptance, and gender biases affect homosexual men’s perpetration of sexual assault. (b) The age at which sexual abuse was first encountered, degree of rape myth acceptance, and gender biases affect the frequency of sexual assault perpetration by homosexual men. (c) According to Ellis (1991) and Seto (2019), targets chosen by sexual assault offenders are determined by their sexual preferences. On the basis of sexual preferences, we hypothesized that men are mostly the target of homosexual men’s sexual assault. (d) If homosexual men’s targets are mostly men, we hypothesized that the influence of male rape myths on homosexual men’s sexual assault is stronger than that of female rape myths. This is because male rape myths comprise a set of prejudices, distorted thinking, or stereotypes against male rape.We followed the research design of Bohner et al. (2005). Two types of questionnaire were developed to collect data online. In Questionnaire B, we placed independent variables (experiences of sexual abuse, two types of rape myths, and gender biases) before dependent variables (sexual assault and frequency of occurrence). The opposite was true for Questionnaire A. Because questions arranged first can trigger participants’ thinking and enhance their cognitive accessibility (heightened temporary accessibility), the stimulation (priming) may affect their answers to the following questions. Therefore, we expected that the association between the independent and dependent variables in the Questionnaire B sample would be higher than that in the Questionnaire A sample, in turn indirectly evidencing the causal relationship between the variables. Additionally, the impact of the order of questions and whether the results were the same in both the samples were also examined. Because Questionnaire A first asked participants about their engagement in sexual assault and frequency, the sensitive nature of the topic might have led to defensiveness; therefore, we predicted that the recognition of rape myth acceptance and gender biases in Questionnaire A would be lower than that in Questionnaire B.We recruited participants who identified as homosexual men and had experiences of undergoing sexual abuse or sexually assaulting others. A total of 414 valid responses were obtained (194 from Questionnaire A and 220 from Questionnaire B). Participants’ background information was as follows: (a) The participants’ average age was 28.28 years, the standard deviation was 6.57 years, the youngest participant was aged 16 years, and the oldest participant was aged 45 years; (b) Educational level: a majority of the participants were at the college level (250, 60.4%), followed by the graduate school level (109, 26.3%); (c) Primary caregiver: parents (307, 74.2%) and grandparents (39, 9.4%); (d) Experience of committing sexual assault: yes (47.3%) and no (52.7%); and (e) Experience of being sexually assaulted: yes (89.4%) and no (10.6%).The t test and chi-square test were used to determine the impact of the three types of risk factors on homosexual men’s sexual assault. Binary logistic regression analysis was performed to examine the impact of the independent variables on the dependent variable, and the amount of variance was explained. Additionally, all the independent variables were placed in the model to examine the effect of the exclusion of overlapping. Pearson correlation analysis and structural equation modeling were used to verify the model on homosexual men’s engagement with sexual assault and its frequency.The results of the study suggested that experiences of sexual abuse, acceptance of rape myths, and gender biases all exert a significant impact and explanatory power on homosexual men’s perpetration of sexual assault. A comparison of the samples of Questionnaires A and B revealed that Questionnaire B had a higher explanatory power for sexual assault perpetrated by homosexual men compared with Questionnaire A (27.1%–36.2% vs. 22.6%–30.3%). This result suggests that regardless of the order of questionnaire items, this study’s hypotheses were supported. Additionally, the result suggested that the independent variables arranged in front of the dependent variables presented a more effective design for confirming our hypotheses compared with vice versa. The average scores of gender biases and rape myths in Questionnaire A were lower than those in Questionnaire B, indicating that sensitive topics presented earlier made participants more cautious or defensive while answering subsequent questions. Furthermore, experiences of sexual abuse, rape myth acceptance, and gender biases influenced not only sexual assault by homosexual men but also their persistence in perpetrating the same. This study confirmed that the majority of homosexual men’s sexual assault targets were men, and regarding myths, male rape myths had more influence on their sexual assaults than did female rape myths.The results of this study revealed that both gender biases and the degree of recognition of rape myths influence sexual assault behavior among homosexual men. Therefore, the sexual aggression behavior signifying the implications of gender bias and power display discussed in past theories can also be observed in homosexual men.The various prevention and publicity measures against sexual assault in our society today are relatively designed from the perspective of viewing women as victims, and myths involving male rape are ignored. However, the impact of male rape myths on sexual assaults by homosexual men is even greater than that exerted by female rape myths, highlighting that male rape myths must be urgently understood and publicized to be eliminated. This may reduce the stigma male victims undergo (McMullen, 1990; Scarce, 1997), which in turn will increase their willingness to report and receive treatment. The perpetrators will have the fear that the victim will come forward, which might help reduce the rate of sexual assault. From a deeper perspective, providing societal acceptance of homosexuality and eliminating myths of male rape will not only reduce the rate of sexual assault, which is extremely underestimated (Coxell et al., 2000), but also aid the addressal of homosexual men’s sexual assault issues.This study recruited homosexual men who had experiences of sexual abuse or sexual assault as participants. Thus, the condition of sexual assault and victimization was widespread. We suggest that future research should include large and representative samples of homosexual men with no restrictions on sexual assault or experiences of being sexually assaulted. This will not only help estimate the prevalence of sexual assault and being sexually assaulted among homosexual men in Taiwan but also further confirm the external validity of the research results.