Objective: Assertive community treatment teams have been criticized as being inherently coercive; however, base rates of control practices used by teams have not been well documented. The purpose of this study was to assess the rates at which different forms of agency control, such as involuntary outpatient commitment, representative payeeship, intensive medication monitoring, and agency-supervised housing, were used by assertive community treatment teams. Also examined were program, practitioner, and consumer correlates of agency control practices. Methods: A statewide survey was conducted of 23 assertive community treatment teams serving consumers with severe mental illness. Data were collected on both team attributes and practitioner attributes. Results: Extent of agency control was highly variable across sites. Intensive medication monitoring and representative payeeship were the most frequently used agency control practices. The strongest predictor of agency control was having a higher percentage of consumers on the caseload who were diagnosed as having a schizophrenia-spectrum disorder. Fidelity to the assertive community treatment model was not associated with agency control. However, lower quality of basic clinical services (for example, assessment and treatment planning) was associated with greater use of agency-supervised housing. Pessimistic practitioner attitudes were not significantly associated with agency control, but practitioner education was negatively associated with both representative payeeship and intensive medication monitoring. Conclusions: Assertive community treatment teams differed widely in their scope of agency control, and this variation was not associated with fidelity to the model. Consumer characteristics, such as a schizophrenia spectrum disorder and active substance use, were most closely associated with agency control. (Psychiatric Services 60:922-928, 2009)