Inhibitors of the cyclooxygenases (COXs), the nonsteroidal antiinflammatory drugs (NSAIDs), relieve inflammatory pain, but are associated with gastrointestinal and cardiovascular complications. Given the widespread use of NSAIDs, there has been a longstanding interest in optimizing their risk-benefit ratio, for example by reducing their gastrointestinal risk. More recently, the focus has shifted toward the cardiovascular complications of NSAIDs and very large prospective studies have been performed to compare cardiovascular risk across distinct NSAIDs. Surprisingly, much less attention has been paid to the efficacy side of the risk-benefit ratio. There is marked variability in the degree of pain relief by NSAIDs due to the complex interplay of molecular mechanisms contributing to the pain sensation, variability in the disposition of NSAIDs, and imprecision in the quantification of human pain. Here we discuss how NSAIDs relieve pain, how molecular mechanisms relate to clinical efficacy, and how this may inform our interpretation of clinical trials.