Characteristics of the posttreatment environment have been implicated in relapse and recovery in substance use disorders. We summarize these findings and describe how different models of relapse interpret the data and diverge in their representations and measurement operations of environmental variables. The advantages and limitations of common measures of the posttreatment environment are discussed using research on life events and relapse as an example. An alternative approach to representing covariation over time between environmental characteristics and substance use that is derived from an application of behavioral theories of choice to relapse is presented.