Theophylline is predominantly metabolized by cytochrome P4501A2 (CYP1A2). A possible interaction between griseofulvin and theophylline was reported to our laboratory, which led us to form the hypothesis that griseofulvin induces the metabolism of theophylline. One purpose of this study was to investigate this hypothesis. The study was carried out as a randomized crossover study of 12 healthy volunteers. In period A of the study, each volunteer received a single dose of 300 mg theophylline ethylenediamine orally. In period B, the subjects took fluvoxamine, 50 mg for 1 day and 100 mg for 6 days, and on day 4, the subjects ingested 300 mg theophylline ethylenediamine. Fluvoxamine is a potent inhibitor of CYP1A2, and period B was included as a positive control. In period C, the subjects took 500 mg griseofulvin for 9 days; on day 8 the subjects again ingested 300 mg theophylline ethylenediamine. Theophylline and its metabolites (1-methyluric acid [IMU], 3-methylxanthine [3MX], and 1,3-dimethyluric acid [13DMU]) in plasma and urine were assayed by high-performance liquid chromatography. During fluvoxamine intake, the median of the total clearance of theophylline decreased from 80 ml/min to 24 ml/min, and the half-life increased from 6.6 to 22 h. The partial formation clearances of the metabolites decreased from 17 to 1.7 ml/min, from 8.9 to 0.9 ml/min, and from 21 to 6.8 ml/min for 1MU, 3MX, and 13DMU, respectively. The results confirm that assessment of theophylline metabolism indeed serves as a biomarker for CYP1A2. During griseofulvin ingestion, the median of the total and partial clearances of theophylline were 84 ml/min, 22 ml/min (1MU), 9.4 ml/min (3MX), and 25 ml/min (13DMU). The half-life decreased significantly from 6.6 to 5.7 h. The increase in partial formation clearances of 1MU and 13DMU, but not of 3MX, were statistically significant. The increase in the total clearance reached only borderline significance. In four subjects a marked induction was seen for all pharmacokinetic parameters, suggesting that the susceptibility to induction is more pronounced in some subjects. This susceptibility could theoretically be explained by a polymorphism in the inducibility of the gene coding for the CYP1A2 enzyme.