Calcium-independent calcium efflux from heavy sarcoplasmic reticulum (HSR) of skeletal muscle was found to be biphasic, with half-times of 2-6 s and 200-400 s for the first and second phase, respectively. Calcium-, AMP- and caffeine-induced calcium efflux was triphasic, with half-times of 0.05-0.2 s, 1-5 s and 100-400 s for the first, second and third phases, respectively. This very fast first phase is certainly due to calcium efflux via the calcium-release channel of HSR vesicles. Both ruthenium red and neomycin inhibited the first phase of the calcium-independent calcium efflux and the first phase of the calcium-, AMP- or caffeine-induced calcium efflux completely, whilst the second phase was fully inhibited by ruthenium red only and partially inhibited by neomycin at high concentrations, indicating that the second phase of calcium release also occurs via the calcium-release channel. Various models for calcium efflux through the release channel have been tested by simulation. Activation and inhibition of the channel-mediated calcium efflux from HSR cannot be explained by two states of the calcium-release channel (open or closed), but requires the existence of at least three states. A channel with one open state and two closed states, resulting in a rapid inactivation, is the most simple model compatible with the experimental data. According to this model, activation is assumed to reduce inactivation of the channel, whilst inhibition assumes an acceleration of channel inactivation. This mechanism most likely applies to neomycin. An additional open-blocked state has to be assumed for inhibition by ruthenium red.