In recent years a growing number of topical nonsteroidal anti-inflammatory drugs (NSAIDs) have become available. This has been prompted in large part by the high incidence of serious gastrointestinal adverse events associated with the use of systemic NSAIDs, and the premise that minimisation of plasma concentrations of active drug may result in fewer systemic adverse effects. Evidence in humans and animals with topical NSAIDs demonstrates lower plasma concentrations than with systemically administered drugs, while those in soft tissues are still of a magnitude considered consistent with exerting an anti-inflammatory effect. In joints, however, the evidence is less strong, and there is still dispute whether in this case the drug reaches the joint predominantly via the transcutaneous or systemic route. There has been a sufficient number of studies of soft tissue conditions to demonstrate the superiority of topical NSAIDs over placebo and to suggest equivalent efficacy in comparison with some oral NSAIDs. For arthropathies, however, the literature is more sparse. Although several studies claim a benefit for topical NSAIDs against placebo, the results are less conclusive and further study is required. Trials of topical agents against intra-articular corticosteroids and rubefacients are either lacking or inconclusive. The adverse event profile of topical agents is reasonable: minor cutaneous effects occur in up to 2% of patients but tend to be self-limiting. Gastrointestinal events appear from the existing literature to be infrequent and minor, although long term studies are required. Bronchospasm and renal impairment have been reported and may be more frequent in patients who have experienced these effects with oral agents. The initial costs of topical agents tend to be higher than those of oral agents but a cost-effectiveness analysis suggests an overall benefit: this issue requires further clarification.