Clinically potent skeletal muscle relaxants are used primarily for their effects on the central nervous system. But they also have direct effects on muscle contraction that possibly involve Ca2+ channels. We compared the effects of dantrolene, an agent known to have a direct action on vertebrate skeletal muscle, with other substances used as (1) relaxants and (2) antagonists of Ca-dependent excitation-contraction coupling. Isolated intact frog muscle cells were injected with the photoprotein aequorin, and membrane potential changes, intracellular Ca2+ transients, and contractile force were measured. Dantrolene (10(-8) to 10(-5) M) decreased the amplitude of Ca2+ transients, did not affect their rates of decay, and reduced contractile force. We also used an integrated digital-imaging system to record microscopic changes, namely, active shortening in myofibrils and changes in striation spacing. Dantrolene did not increase the time between contraction in myofibrils near the surface compared with myofibrils near the center of a cell. Hence dantrolene does not suppress Ca2+ transients by disturbing current flow in the transverse tubular system. Each of the following actually increased Ca2+ transients and contractile force evoked by action potentials: baclofen (10(-7) to 10(-5) M) less than flordipine (10(-6) M) less than meprobamate (10(-7) to 10(-3) M) less than chlordiazepoxide (10(-5) X 10(-4) M) less than procaine (10(-5) to 5 X 10(-4] less than GABA (10(-5) M) less than D-600 (10(-6) M) less than nylidrin (10(-5) M)--in order of increasing potency. Ca2+ channels in the sarcoplasmic reticulum of intact skeletal muscle are evidently inhibited by dantrolene but not by Ca2+ antagonists.