Specific inhibitors of cyclooxygenase-2 were introduced into widespread clinical use in 1999. Since that time, celecoxib and rofecoxib have become two of the most commonly prescribed medications in the United States. Clinical trials using these medications for arthritis and pain have uniformly demonstrated efficacy superior to that of placebo and similar to that of nonsteroidal anti-inflammatory drugs. However, controversy surrounding the proper place of cyclooxygenase-2 inhibitors in the hierarchy of treatment for arthritis continues, based primarily on their higher cost compared with that of acetaminophen and nonsteroidal anti-inflammatory drugs. A decreased risk of gastrointestinal toxicity remains the primary justification for using the more expensive cyclooxygenase-2 inhibitors in preference to nonsteroidal anti-inflammatory drugs. The renal and cardiovascular effects of rofecoxib and celecoxib have been investigated in relation to nonsteroidal anti-inflammatory drugs and to one another. The data with respect to alteration in renal function, lower extremity edema, and hypertension indicates that cyclooxygenase-2 inhibitors affect the kidney in a manner similar to that of nonsteroidal anti-inflammatory drugs. The data comparing the cyclooxygenase-2 inhibitors is difficult to interpret because it is not clear that comparable doses have been used in clinical trials. The potential thrombogenic risk of cyclooxygenase-2 inhibitors remains controversial, and conflicting data exist. It remains important to increase our understanding of the place of these agents in clinical practice from the perspective of efficacy, toxicity, and cost.