Sexual homicide is a rare occurrence. Little is known about the offending perspective of sexual homicide from a criminological standpoint. Recently, Chan, Heide, and Beauregard (2011) proposed an integrative theoretical framework using concepts and propositions of Social Learning Theory (differential association, definitions, differential reinforcement or punishment, and imitation) and Routine Activities Theory (a motivated offender, an attractive and suitable target, and the absence of a capable guardian or guardianship) to elucidate the sexual homicide offending dynamics. According to this integrative model, the individual-level view of the sexual murderers is explained by the social learning principles, while the offending process is complemented by the routine activities propositions from a micro-level to provide a better explained sexual homicide offending model. However, this model has yet to be tested empirically. In addition to testing the Chan et al.'s model, this study proposes and tests an alternative model by incorporating the construct of pre-crime precipitators to better explain the motivating factor of an offender to commit a sexual homicide. To empirically test both models, this study utilizes the dataset collected by a group of Canadian researchers on 230 incarcerated non-serial homicidal (N = 55) and non-homicidal (N = 175) sex offenders in the province of Quebec, Canada for the period between 1995 and 2005. Using step-wise logistic regression, four regression models are tested to examine the offending process of sexual homicide by investigating the effects of the offender's motivation, the target suitability and attractiveness, the absence of a capable guardian or guardianship, and the pre-crime precipitating factors in deciding the lethal outcome of a sexual offense. The theoretical model proposed by Chan and colleagues received some support. Consistent with Chan et al.'s theoretical propositions, findings suggest that the sex offender's sexually deviant behaviors and attitudes serve as a motivating factor, and the presence/absence of a capable guardian or guardianship at the immediate crime surroundings are significant factors in deciding the survival rate of the victim. Methodological limitations of the study, practical implications for offender profiling and crime preventive measures, and suggestions for future research are discussed.