Differences in the personal nature of police officers as well as differences among the situational constraints due to assignments are considered in an examination of the occurrences of police involvement in shooting incidents. The Poisson and negative binomial models used in accident research are considered for representing the occurrence of shooting incidents. Data were gathered from incidents in a large metropolitan police department over a 34‐month period and used to test the two models. Various analyses indicate a good Poisson fit. Implications of the data are discusse