OBJECTIVES: To understand the patterns of aspirin utilization among users of Nonsteroidal Anti-Inflammatory Drugs (NSAIDs) with varying cardiovascular risk profiles. METHODS: A telephone survey was completed using a large health benefits company population. The survey consisted of 10-items used to characterize aspirin consumption and identify motivation for utilization. Subjects were randomly selected from a dataset of members aged 18 and older, who had an NSAID prescription claim between April 1, 2003 and June 30, 2003, while maintaining continuous plan enrollment during a 24-month period. Subjects were also required to maintain chronic NSAID utilization, defined as at least a 90 days' supply during a 12-month period. Study subjects were stratified based upon their NSAID utilization into: 1) cox-II selective inhibitors; 2) non-selective NSAIDs (excluding naproxen); and 3) naproxen. Subjects were further stratified based upon the presence or absence of risk factors associated with cardiovascular events. RESULTS: The study population consisted of 1250 subjects, of which 52.3% were treated with non-selective NSAIDs, 19.4% with naproxen, and 28.3% with cox-II inhibitors. In total, 77.1% of the population had risk factors associated with cardiovascular events. No significant difference (P=0.6945) was found in the proportion of subjects using aspirin among the non-selective NSAID, naproxen, and cox-II cohorts, with 46.8%, 48.8%, and 49.4%, respectively. Likewise, no significant differences were found among the treatment cohorts with respect to: strength, frequency, and duration of aspirin use (P=0.3840, P=0.8088 and P=0.6838, respectively). Finally, no significant difference (P=0.2778) was found in the proportion of subjects using aspirin among those with risk factors for cardiovascular events versus those without. CONCLUSIONS: Unexpectedly, these results indicate that aspirin utilization, strength, frequency, and duration are independent of both subjects' cardiovascular risk profile (i.e. risk vs. no risk) and the NSAID class utilized (i.e. selective vs. nonselective).