1. We studied the effects of caffeine on coronary artery smooth muscle of the pig by measuring changes in isometric tension, cytosolic free Ca(2+) concentration ( [Ca2+]i) and transmembrane potential. 2. In the absence of tone, caffeine induced a concentration-dependent transient contraction of coronary artery strips, followed by sustained relaxation. Simultaneously with the relaxation, caffeine, 25 mM, hyperpolarized the smooth muscle cells by 7.7 +/- 0.9 mV. 3. Caffeine caused a concentration-dependent relaxation of strips precontracted with 10(-5)M acetylcholine (ACH). A supramaximal relaxing concentration of 25 mM caffeine produced an additional transient increase in [Ca2+]i on the Ca2+ plateau of ACh tonic contraction, which was followed by a decrease in [Ca2+]i to a level slightly below the basal concentration. This relaxation was accompanied by a hyperpolarization of 7.3 +/- 0.9 mV. 4. KCI 120 mM (high K+) contracted the strips with a concomitant depolarization of 38.6 +/- 1.6 mV and sustained increase in [Ca2+]i. Caffeine caused a concentration-dependent relaxation of high K+-induced contraction. Caffeine, 25 mM, decreased the Ca2+ plateau to a level that remained above the basal concentration of Ca2+ but did not change the membrane potential. 5. When strips were placed in a Ca(2+)-free medium with EGTA 2mM, and, in addition, ACh was applied successively three times, both intracellular and extracellular mobilizable Ca2+ pools were depleted. In these conditions, phorbol 12,13 dibutyrate (PDBu) 10(-7) M and prostaglandin F 2 alpha (PGF 2 alpha) 10(-5) M contracted the strips. Caffeine (25 mM) inhibited these contractions with no change in [Ca2+]i. 6. Forskolin, 3 x 10 -7M, inhibited ACh induced-contraction but did not affect those induced by PDBu. 7. In conclusion, these results show that caffeine has multiple cellular effects. During caffeine-induced relaxation, [Ca2" Ii, adenosine 3': 5'-cyclic monophosphate (cyclic AMP) content and membrane potential are modified. The findings suggest, however, that these effects are secondary, and that caffeine acts mainly by another unknown mechanism, possibly involving a direct inhibition of the contractile apparatus.