An empirical perspective is presented illustrating the application of the principles and methods of community psychology for conceptualizing the ideal of "social justice." The substantiative content of the paper is based on an analysis of the selective nature of sexual assaults which reach the criminal justice system, and in the disparity of sentences given to those found guilty, in comparison to other types of cases. These outcomes are proposed as operational definitions for conceptualizing the ideal of social justice. The application of this standard for cases of sexual assault suggests the justice system "discounts" the seriousness of these offenses, resulting in a legal doctrine which excuses male violence against women. Implications for law reform are proposed based on the need for an accommodation of the social/psychological concept of "difference" in the legal definition of sexual assault, and in the greater use of evaluation research as a process of legal scholarship for monitoring how well the ideal of social justice is achieved by the criminal justice system.One purpose of the present paper is to highlight the value of using a community psychology framework for conceptualizing important social issues that adversely affect the human condition, and of the responsibility of psychologists to address these issues through empirical scholarship. The methodological perspective which follows is based on a research program that began over a decade ago and now includes three principal data sources: The first is based on the records of an agency which provides services for victims of sexual assault. This data, supplemented by two national victimization surveys (Solicitor General of Canada, 1983, 1984; and Statistics Canada, 1993), provide the best available description of the true nature and frequency of incidents of sexual assault. The second source, the court records, provides a formal account of the outcome of the criminal justice process. The third data source, comprised of transcripts of court cases, provides a qualitative reference point that ensures that the quantitative analyses of the court records are capturing the essential aspects of actual courtroom dynamics, i.e., that the statistical analysis has ecological validity.Two of the three methodologies have been previously described in the literature and will not be re-described in detail here. Briefly, the case data now includes 2,533 consecutive cases responded to by a sexual assault service. (For methodological details see Renner & Wackett, 1987; Renner, Wackett & Ganderton, 1988 for previous reports based on the initial 1,236 cases.) The court record data now includes a run of 1,074 consecutive court cases tried in the same city, composed of sexual assaults (n=354), and, for comparative purposes, physical assaults ([Symbol Not Transcribed] = 513) and robberies ([Symbol Not Transcribed] = 207). (For methodological details see Yurchesyn, Keith & Renner, 1992; Renner & Yurchesyn, 1994 for previous reports based on the initial 356 cases of the now larger sample.) These two expanded data sources have been supplemented by a sample of 104 transcripts of 51 adult and 53 child sexual assault cases from the court records data files. This third data source was the primary basis of the qualitative propositions to be advanced and evaluated in this paper(f.1).These three data sources converge to provide a descriptive account of how the crime of sexual assault on women and children by men is addressed by the formal legal doctrines which have been put into place by the criminal justice system. Our research does not address the frequency or nature of sexual offenses by women against men or children. Thus, the second purpose of this paper is to examine how cases of male sexual violence are handled by the criminal justice system from the perspective of whether the standard of "social justice" is being met. Social justice is the ideal of seeking the truth in an effort to provide fair and just outcomes through the criminal justice process (Renner & Warner, 1981). …