Six randomized, double-blind, two-period crossover studies, conducted under similar protocols, compared the efficacy of two analgesic combinations containing caffeine with an acetaminophen 1000 mg control and with a placebo in outpatients with episodic tension-type headaches. In four studies, comprising 1900 patients, the caffeine-containing analgesic consisted of a combination of 500 mg acetaminophen, 500 mg aspirin, and 130 mg caffeine (APAP/ASA/CAF). In two studies, comprising 911 patients, the caffeine-containing analgesic consisted of a combination of 1000 mg acetaminophen and 130 mg caffeine (APAP/CAF). Patients self-medicated for moderate or severe headache pain, and with a self-rating record they rated their pain and its relief hourly for 4 hours. In all six studies, the caffeine-containing analgesics were significantly superior both to placebo and to 1000 mg acetaminophen, and acetaminophen was significantly superior to placebo. The significant analgesic adjuvant effect of caffeine was independent of patients' usual caffeine use or their caffeine consumption in the 4 hours before medication. For each treatment, the pooled analgesic responses for the four studies of APAP/ASA/CAF were virtually superimposable on the responses in the two APAP/CAF studies. The combinations produced more stomach discomfort, nervousness, and dizziness than acetaminophen or placebo.