The suction blister technique was used for pharmacokinetic studies with sulfonamides and trimethoprim. Blisters produced by suction (-0.3 kg/cm2) for 1.5 h contained approximately 0.15 ml fluid with a protein content of 40-50% of that in plasma, the main protein fractions being present in the same ratio as in plasma. 2 g sulfaisodimidine was given as bolus injection, i.v. infusion or orally to groups of 4 volunteers. The peak blister fluid concentrations after oral administration (120 +/- 18 mmol/l) was only marginally lower than the concentrations after i.v. infusion (122 +/- 28 mmol/l) and i.v. bolus injections (134 +/- 37 mmol/l). The total drug blister fluid concentration started to decrease before the plasma level was reached. However the relative concentration increased from 53% of that in plasma at 8 h to 66% at 12 h after drug administration. Considering the protein binding of the drugs, the interstitial fluid levels of free drug were presumably higher than the plasma level after 8 h. Comparison of drug concentrations in blisters produced before and after the drugs were given showed higher concentrations in the latter for the first 2-6 h. However, after 8-12 h the concentrations of the drugs in the two types of blisters were similar. The suction blister method produces blisters of uniform size. The drug concentrations in different experiments showed the coefficient of variation for blister fluid concentrations to be no greater than for plasma levels. The consistent results of the standardized suction blister method makes this method useful for studying drug penetration to extravascular compartments in humans.