BACKGROUND AND OBJECTIVE: Pseudoallergic reactions triggered by nonsteroidal anti-inflammatory drugs (NSAIDs) are common and caused by inhibition of the enzyme cyclooxygenase-1, whereas their therapeutic effects are mediated by inhibition of cyclooxygenase-2. This study analyzed the tolerability of the selective cyclooxygenase-2-inhibitor rofecoxib in patients who encountered pseudoallergic reactions to NSAIDs. PATIENTS AND METHODS: 37 patients (12 males, 25 females, mean age 35 years [15-75 years]) with a history of pseudoallergic reactions to NSAIDs underwent standardized skin prick, scratch and patch tests along with oral placebo-controlled blinded exposure to rofecoxib (maximum single dose 12.5 mg, cumulative dose 25 mg). RESULTS: 23 patients had skin reactions, 4 times respiratory symptoms were documented, and in 10 cases cutaneous as well as respiratory symptoms were reported. Salicylic acid was identified as the most common trigger for a pseudoallergic reaction (n = 28). In 9 cases several non-steroidal antiphlogistics of different chemical groups caused symptoms. All skin tests showed negative results. Oral challenge with rofecoxib was tolerated by all 37 patients without adverse effects. CONCLUSION: Given the high incidence of pseudoallergic reactions to NSAIDs the use of selective cyclooxygenase-2-inhibitors represents a therapeutic alternative as well as a means of prevention of the described reactions.