The authors sought to determine whether the metabolites of neostigmine and pyridostigmine contribute to antagonism of neuromuscular blockade. Accordingly, the dose-response relationship, onset and duration of action (n = 60), and pharmacokinetics (n = 22) of neostigmine, pyridostigmine, their metabolites 3-hydroxyphenyltrimethylammonium (PTMA) and 3-hydroxy-N-methylpyridinium (MP), and edrophonium were determined in dogs anesthetized with sodium pentobarbital. The force of contraction of the anterior tibialis muscle was maintained at constant 90% depression by infusing pancuronium. Then, a single iv bolus dose of one of the drugs under study was injected while the pancuronium infusion was continued. Venous blood, urine, and bile were sampled for four hours. Concentrations were determined by liquid chromatographic techniques; a three-compartment pharmacokinetic model was fitted to the serum concentration data. The doses producing 50% antagonism were 6.5, 52, 69, and 40 micrograms/kg for neostigmine, pyridostigmine, edrophonium, and PTMA, respectively. MP was inactive as an antagonist. By comparing approximately equipotent doses, time to peak antagonism (onset) and until 30% of peak antagonism remained (duration) were shorter for both edrophonium and PTMA than for neostigmine and pyridostigmine. Slow distribution and elimination half-lives, volume of distribution at steady state (VDss), and total plasma clearance (Cl) were similar for the drugs except for a smaller Vdss and lower Cl for MP. More than 60% of the dose of each drug was recovered unchanged from urine; less than 1% was recovered from bile. Less than 10% of the dose of neostigmine was recovered as PTMA.(ABSTRACT TRUNCATED AT 250 WORDS)