The objectives of this study were to determine the relationship between drug therapy compliance and risk of hospitalization and economic outcomes, and to identify potential indicators of compliance. We used computerized prescription records from 1,054 patients at high risk for drug-related problems. We calculated a compliance ratio for a 12-month period and correlated it with health care use, demographic variables, drug-related variables, and scores for health-related quality of life. Univariate results suggested that increased age (p=0.05), high number of chronic conditions (p<0.001), and high number of concurrent drugs (p<0.001) were positively correlated with compliance. That is, increased values for these variables were associated with better compliance. Using logistic regression, the odds of being noncompliant was 0.665 as the number of chronic conditions increased. Compliance was not a predictor of concurrent or future hospitalizations or mortality, nor was it a significant predictor of health care costs.