This study is a methodological improvement from previous cross-national homicide studies. First, relatively large numbers of nations were included in the model. Ninety-two countries from the World Health Organization (WHO) and 90 nations from the International Police organization (Interpol) have contributed data. Second, the current model introduced many variables compared with previous studies, as it used 13 variables, including 7 medical variables, 4 variables derived from competing theories, and 2 other variables. Further, the study used an additional dependent variable, lethality rate, which refers to the ratio of successful to attempted homicides. Finally, the analysis used distinct efforts to minimize the weaknesses typically associated with international homicide data, as it refined the Interpol data adjusted by attempted homicides, thus showing only "true" or "actual" homicides. In addition, averages of multiple year homicide rates between 1990 and 1998, which included the most recent available data, were used to relax the problem with erratic data reporting. The study used multiple regression techniques to test the relationship between medical resources and homicide rates across nations while controlling variables for competing theories such as gross national product per capita, income inequality, ethnic heterogeneity, and dummy variables for geographical regions. The study found only weak support for the proposed medical model; however, it is premature to discard the medical perspective as a whole. As discussed, there were a number of intrusive limitations associated with the current study. Future researchers must address those limitations before reaching a definitive assessment as to whether medical variables influence the production of criminal lethality. 20 tables, 6 figures, 224 references, and appended lists of study groups for dependent variables