BACKGROUND: Dantrolene produces skeletal muscle relaxation by a direct action on excitation-contraction coupling, presumably by decreasing the amount of calcium released from the sarcoplasmic reticulum. The mechanism underlying this action is extrajunctional. The aim of this study was to evaluate the pharmacodynamic properties of dantrolene at the neuromuscular junction and the reversal effects of substances as possible dantrolene antagonists in vitro. METHODS: The effects of evoked twitch tension response have been studied on the isolated phrenic nerve hemidiaphragm muscle strips of the rat, using a single twitch (0.1 Hz) and the train of four (TOF; 2 Hz for 2s) stimulation. The maximum effect (E(max)) and TOF ratio at each point of twitch depression after cumulative doses of dantrolene were measured mechanomyographically. The EC(50) and EC(95) of dantrolene were calculated using an inhibitory sigmoid E(max) model. The reversal effect to E(max) after administration of 10 mM of dantrolene was determined by various doses of neostigmine, pyridostigmine or 4-aminopyridine respectively. RESULTS: The E(max) was 76.14% of the initial twitch tension, but the residual twitch tension was remained until five times (10 mM) of the dose for the E(max) was administered. TOF stimulation to the residual twitch tension did not demonstrate any fade. The EC(50) and EC(95) of dantrolene were 0.379 and 3.177 mM respectively. Neostigmine and pyridostigmine produced a transient but incomplete recovery of twitch tension, which rapidly fell to the level of the twitch response before the drugs were given. However, 4-aminopyridine produced a dose-dependent recovery of the twitch response. The addition of neostigmine (0.5 mg/ml) or pyridostigmine (2.5 mg/ml) did not decrease the EC(50) and EC(95) of 4-aminopyridine in reversing the effect of dantrolene. CONCLUSIONS: These RESULTS have demonstrated the evidence that dantrolene did not completely depress the twitch tension, leaving if at nearly 25%, and accompanying TOF response without fade, and that anticholinesterases were ineffective in antagonizing its blockade. However, 4-aminopyridine was effective and may not be related to the propensity for pre- and postjunctional cholinergic receptor blockade at the neuromuscular junction.