Multiple murder encompasses a heterogenous group of acts including spree killings, mass and serial murders, and sexually motivated murders. Of 11 studies of multiple murder examined, 9 were either unevaluated case reports or scientific case reports of individuals or small groups. All were retrospective. Because such studies are not generalizable to the entire population, their underlying causal assumptions (e.g., psychiatric disturbance, life history factors) cannot be proven. Further, small sample size combined with retrospection introduces a bias that tends to confirm the researcher's assumptions. Only one study (Ressler et al, 1984) provides an example of an epidemiologic study, but, again, use of retrospective interviews may have introduced biases, and the information yielded is not sufficiently predictive or explanatory. Thus, none of these studies supports a primary hypothesis that explains the data either in psychiatric or criminologic terms. The literature does, however, provide a basis for formulating testable hypotheses. Examining the way in which personal and constitutional factors interact with contextual variables in large-scale research projects will aid in identifying the important causal factors in the etiologies of this complex behavioral syndrome. 50 references, 2 figures, and l table.