OBJECTIVES: Traditional nonsteroidal anti-inflammatory drug (tNSAID) utilisation has recently increased considerably in the elderly in spite of their higher gastrointestinal (GI) toxicity and an increasing body of evidence pointing to similar cardiovascular (CV) safety compared to celecoxib. The objective is to describe the profile of patients who used celecoxib and tNSAIDs between April 1, 2005 and March 31, 2007 (post- rofecoxib withdrawal period) versus April 1, 2002 and March 31, 2004 (pre-rofecoxib withdrawal period) in Quebec, Canada. METHODS: Data were obtained from the physician and medication claims databases of the Quebec Health Insurance Agency. All patients 50 years of age and older who used celecoxib or tNSAIDs during the study periods were included. Patients were categorized by GI, CV, congestive heart failure (CHF) and renal risk factors and four risk categories were considered (low, moderate, high, and very high) in each condition. RESULTS: The numbers of patients on celecoxib were 145,596 and 178,714 in the post- vs. pre-period, while those on tNSAIDs were 249,433 and 120,809, respectively. Logistic regression models revealed that users of celecoxib vs. tNSAIDs were more likely to have higher GI risk levels in both periods [post-period (vs. low risk): very high 1.79 (95%CI: 1.63, 1.97), high 1.76 (1.71, 1.81) and moderate 1.30 (1.27, 1.33)]. While users of celecoxib had higher CV risk levels in the pre-period compared to tNSAIDs, their very high CV risk level in the post-period was lower [post-period (vs. low risk): very high 0.85 (0.81, 0.89), high 1.13 (1.10, 1.16) and moderate 1.15 (1.12, 1.17)]. In both periods, the impact of CHF and renal risks on the choice of celecoxib vs. tNSAIDs appears to be less important. CONCLUSIONS: Currently, patients at very high CV risk seem to be less likely to receive celecoxib vs. tNSAIDs, while those with GI risk factors seem more likely to receive celecoxib.