The Canadian data covered the years 1961-83 and provided a data set of 9,642 cases. The United States data covered 1976-92 and produced a data set of 341,369 cases. The data analysis used logistic regression techniques to predict homicide clearance on the basis of specific victim and offense characteristics for cases in Canada versus the United States and in Ontario, Canada, versus New York State. Results indicated that the model was a good fit for homicide clearance in both countries as a whole. In addition, the homicide weapon, the circumstances surrounding the offense, victim’s age, and victim’s gender were significant predictors of homicide clearance in New York State. In contrast, only the circumstances surrounding the offense emerged as an important predictor in Ontario. Tables, notes, and 48 references (Author abstract modified)