Several non-steroidal anti-inflammatory drugs (NSAIDs) have been associated with liver damage. The aim of this study was to compare proportions of hepatic adverse drug reaction reports associated with NSAIDs in France and Spain. Information from the Spanish and French pharmacovigilance databases were used from 1982 to 2001. To assess the risk of liver injury, the case/non-case methodology was applied, 'cases' being reports of liver damage and 'non-cases' or controls, all other reports. Exposure was considered as the presence of at least one NSAID. Liver injury risk was estimated for each drug in the two databases by calculation of reporting odds ratio in cases and non-cases, with its 95% confidence interval. Out of 62 456 reports from the Spanish database, 2114 (3.38%) were identified as liver injuries, whereas there were 27 372 liver injuries out of 200 046 (13.68%) in the French database. In Spain, there was a significant association between liver injuries and droxicam, sulindac, and nimesulide. The risk was also slightly above 1 for aceclofenac. In France, the risk was very high with clometacin, followed by sulindac, and was slightly above 1 for naproxen, diclofenac, piroxicam, and tenoxicam. This study shows that some NSAIDs are associated with reports of hepatic injuries when compared with other drugs, and most of those have been withdrawn from the market for this reason. However, the frequency of drug-related hepatic injuries reported differed in the French and Spanish databases, and some drugs did not show the same risk level in the two countries. These discrepancies could be explained in part not only by reporting rates, but also by difference in drug use patterns and/or by genetic or environmental factors.