A retrospective cohort study using 1980-1984 Medicaid billing data from 3 US states was performed to assess the relative risk of hypersensitivity reactions from different nonsteroidal antiinflammatory drugs (NSAID). Comparing tolmetin sodium to other NSAID, pooling the data across the 3 US states using the Mantel-Haenszel procedure, gave an overall relative risk (95% confidence interval) of 1.1 (0.8-1.4). After adjusting for multiple potential confounding variables using logistic regression, an odds ratio (95% confidence interval) of 0.9 (0.6-1.2) was observed. These data do not confirm previous suggestions that use of tolmetin is associated with a higher risk of hypersensitivity reactions than use of other NSAID.