This study is one of the few to investigate correlates of force in the Canadian context. It also investigates the existence of protective factors that decrease the level of force used by the police. A total of 1,174 self-reported uses of force are analyzed. Multinomial logistic regression models were used to identify factors related to three possibilities: The force used by the police was lower than, equal to, or higher than the level of subject resistance. The analysis reveals that the impact of individual characteristics on the correspondence between officer force and subject resistance is negligible. Also, three general patterns of relationships are found. First, the presence of a weapon helps distinguish lower-than-expected force situations. Second, the presence of a single officer, resistance toward officer(s), conflict between the subject and another citizen, and subject intoxication have linear effects, that is, the effect increases or decreases consistently. Third, for every less severe level of force that was used, cases are more likely to be in the expected than the lower-than- and in the higher-than-expected group. The findings obtained in this study are consistent with the literature, suggesting that it is reasonable to apply most conclusions from previous studies on police use of force to the Canadian context. The analysis also suggests that police use of force could be better understood as a trichotomy where the force used by the police is depicted as lower than, equal to, or higher than the level of subject resistance.