The criminalization hypothesis is based on the assumption that police inappropriately use arrest to resolve encounters with mentally disordered suspects. The current study uses data collected from two large‐scale, multisite field studies of police behavior‐the Project on Policing Neighborhoods (POPN) conducted in 1996–1997 and the Police Services Study (PSS) conducted in 1977‐to examine the relationship between suspect mental health and use of arrest by police. Multivariate results show that police are not more likely to arrest mentally disordered suspects. Implications for future research on the criminalization hypothesis are discussed.