Enzymatic hydrolysis of acetylcholine (ACh) was determined in intact frog sartorius muscles or their homogenates. The Vmax was 29 nmol min-1 in intact muscles and 46 nmol min-1 per muscle in homogenates, and the Km was 6 and 0.2 mM, respectively. The muscle was divided into small segments, which were homogenized; the junctional cholinesterase (ChE) accounted for 60% of total enzyme activity. At low substrate concentrations the rate of hydrolysis was up to 30 times higher in homogenates than in intact muscles. This difference was greatly reduced at very high substrate concentrations. It appears that most of the ChE in intact muscle is 'occluded' to external ACh, mainly because the ChE at the edges of the synaptic cleft prevents the ACh from reaching the enzyme situated further inwards, which consequently does not contribute to its hydrolysis; homogenization makes all synaptic ChE accessible to added ACh. Incubation of sartorius muscles with collagenase caused an 80% decrease in ChE activity (determined in homogenates) of end-plate-containing parts which became similar to that in end-plate-free parts on which collagenase had little effect. Histochemistry showed that the tendon-muscle junction contained folds which were stained intensively for ChE. Diethyldimethylpyrophosphonate , neostigmine, eserine, and di-isopropyl fluorophosphonate inhibited ChE activity in this order of potency. The I50 values (i.e. the concentrations of the drugs which caused a 50% inhibition) were about 5 times higher in intact than in homogenized tissue. Neostigmine, 0.15 and 0.4 microM, increased the time constant of miniature end-plate currents 1.3- and 1.8-fold, and slowed down ChE activity of muscle homogenates by 1.4 and 2.1 times, respectively, without significantly affecting ACh hydrolysis by intact muscles. This indicates that synaptic ChE is not present in large excess. It is concluded that ChE activity measured in homogenates presents a better picture of in situ ChE activity than that measured in whole muscles especially for evaluating the effect of ChE inhibitors. A mathematical model for ChE-hindered diffusion of ACh is presented in an Appendix.