1. The effects of cocaine on the neuromuscular transmission and smooth muscle cells of the guinea-pig mesenteric artery were observed using various experimental procedures.2. Cocaine (10(-7) M) depolarized the membrane and increased the membrane resistance of single smooth muscle cells. Outward current pulses produced neither spikes nor graded responses in Krebs solution, but in the presence of 10(-5) M-cocaine, outward current did produce spikes.3. Perivascular nerve stimulation evoked excitatory junction potentials (e.j.p.s). Repetitive stimulation (0.25-1.0 Hz) produced a frequency-dependent facilitation. Application of cocaine (10(-7) M) reduced the amplitude of the first e.j.p. (e.j.p.(f)) and also after completion of facilitation (e.j.p.(s)). However, the facilitation process was not affected by cocaine (10(-5) M).4. On pre-treatment with phentolamine (3 x 10(-7) M), both e.j.p.(f) and e.j.p.(s) were enhanced, but on pre-treatment with yohimbine (3 x 10(-7) M), e.j.p.(f) was inhibited and e.j.p.(s) was enhanced. Both phentolamine and yohimbine accelerated the facilitation of e.j.p.s in the absence and presence of cocaine (10(-7) M).5. The conduction velocity of nerve excitation measured from the latency of generation of e.j.p.s was slightly lowered by cocaine. The number of nerve fibres or varicosities contributing to the generation of an e.j.p. was not reduced in the presence of cocaine (10(-5) M).6. Mechanical responses could be recorded on perivascular nerve stimulation, and direct muscle stimulation on treatment with tetrodotoxin. Cocaine (10(-7) to 10(-4) M) enhanced the contraction evoked by direct muscle stimulation and inhibited the contraction evoked by perivascular nerve stimulation.7. Cocaine (10(-5) M) enhanced the contraction evoked by 5 x 10(-6) M-noradrenaline (NA) and direct muscle stimulation (5 sec pulse) but no effect was observed on the K-induced contraction (39.2 mM-K). On pre-treatment with guanethidine (10(-6) M) these effects of cocaine were not affected.8. In the presence of cocaine (10(-5) M), the depolarization of the membrane induced by NA was additively increased, and the dose response curve for NA was shifted to the left with no change in the maximum amplitude of contraction.9. When 10(-5) M-cocaine was applied during contractions evoked by alternate perivascular nerve stimulation and exogenously applied NA, the contraction evoked by perivascular nerve stimulation was reduced, while that evoked by NA was enhanced.10. In saponin-treated skinned muscles, the pCa-tension relationship was not affected by application of 10(-4) M-cocaine. The effects of cocaine on the Ca accumulation and release from the store site were estimated. It was found that cocaine (10(-4) M) slightly inhibited the Ca accumulation (0.89 times the control) but did not modify the Ca-release mechanism.11. The overflow of NA, 3,4-dihydroxyphenylglycol (DOPEG) and 3-methoxy-4-hydroxyphenylglycol (MOPEG) were measured in the same tissue before and after application of perivascular nerve stimulation in the presence or absence of cocaine (10(-7) to 10(-5) M). Cocaine induced a concentration-dependent increase in the overflow of NA and a reduction in the amounts of DOPEG and MOPEG.12. We conclude from these studies that cocaine mainly inhibits the sensitivity of the intra-junctional adrenoceptor, but increases the sensitivity of the extrajunctional adrenoceptor distributed on the post-junctional muscle membrane, with increase in the overflow of NA. The enhancement of mechanical response in the presence of cocaine is probably due to an increased sensitivity of the extra-junctional adrenoceptor and changes in the post-junctional muscle membrane, without any marked effect on the prejunctional mechanism.