403 patients suffering from moderate pain associated with osteoarthritis (OA) of the knee were randomized into three groups. The groups consisted of an acetaminophen extended-release (APAP ER) group, a 12.5 mg rofecoxib group, and a 25 mg rofecoxib group. Following four weeks of treatment each group produced a significant reduction in WOMAC pain scores from baseline measures. APAP ER was found to be non-inferior, in terms of reduction in pain scores, when compared to the 12.5mg dose of rofecoxib. However, this non-inferiority was not established in comparison to the larger rofecoxib dose.