Oxidative drug metabolizing capacity has been assessed by oral antipyrine and/or theophylline tests in consecutive patients with chronic pancreatitis (CP; alcoholic 24, idiopathic 47), acute pancreatitis (AP; 28) and pancreatic cancer (CA; 11). Most of the patients had drastically reduced their alcohol consumption and dietary fat intake for variable periods before the tests. Excellent bioavailability of theophylline was confirmed from paired oral and intravenous tests in seven subjects, including two with exocrine pancreatic failure. 2. The clearance of theophylline in the patients was faster than in 15 controls with a 'healthy lifestyle' [median 104 (range 18-320) ml h-1 kg-1 vs median 68 (range 50-97) ml h-1 kg-1, P less than 0.01]. The difference was especially apparent in the groups with alcoholic CP (P less than 0.001 and idiopathic CP (P less than 0.01), but not in the AP and CA group as a whole, although drug clearance in some 50% of those cases exceeded the reference range. 3. There was good correlation between theophylline and antipyrine clearance in a subset of 91 subjects who had both tests (15 controls, 76 patients), but antipyrine was much less sensitive as a marker of enzyme induction. This suggests that enzyme induction in pancreatic disease preferentially involves the polycyclic aromatic hydrocarbon-inducible subfamily of cytochrome P-450. 4. The lack of correlation between pancreatic secretory capacity in 56 cases, judged by a secretin-pancreozymin test, and theophylline clearance suggests that enzyme induction is not secondary to pancreatic dysfunction. 5. Multivariate regression analysis identified approximately 50% of variability in clearance of each probe.(ABSTRACT TRUNCATED AT 250 WORDS)