A matched-pair case-control analysis of Medicaid claims was performed to determine the risks of hospitalization associated with drug-drug interactions. Patients were hospitalized and controls were not. They were randomly matched based on contemporaneous eligibility for Medicaid benefits. Odds ratios for hospitalization in patients exposed to known drug-drug interactions were compared with those in patients exposed to one of the interacting agents. When confidence intervals did not overlap, the odds ratio was considered to be significantly increased. Odds ratios were significantly increased for many interacting drug pairs, and were associated with commonly recognized interactions as well as less widely recognized ones. Cimetidine interactions achieved significance only with theophylline. In the Medicaid population, exposure to a number of drug-drug interactions was associated with a significantly increased risk of hospitalization.