It has long been assumed that there exists a relationship between crime (including homicide) and season. After discussing three analytic approaches to this problem (looking for seasonality, a general autoregressive process, and cycles), we review the literature and show that confusing and conflicting findings have been reported about the temporal regularities of homicide. Employing monthly data from the Supplementary Homicide Reports (1976–1989), we find evidence for seasonality, autoregression, and cyclicality of homicide. Our modeling approaches clarify the previous conflicting research; implications for theory and future research are discussed.