The effect of convulsive stimulations on the dissipation of tolerance to the anticonvulsant effect of diazepam was investigated using the kindled-convulsion model. Amygdala-kindled rats were rendered tolerant to diazepam's anticonvulsant effect by 25 "bidaily" (one/48 h) diazepam injections (2.5 mg/kg), each followed 1 h later by a convulsive stimulation. They were then divided into nine groups for the tolerance-dissipation phase of the experiment. Of the nine groups, three received bidaily control handling for one trial, three trials, or seven trials; three received bidaily saline injections, each 1 h before a convulsive stimulation, for one, three, or seven trials; and three received bidaily diazepam injections, each 1 h after a convulsive stimulation, for one, three, or seven trials. Finally, each rat received a tolerance-retention test (i.e., a diazepam injection followed 1 h later by a convulsive stimulation) 48 h after its last tolerance-dissipation trial. The tolerance dissipated gradually but completely over the 4-, 8-, and 16-day test intervals in the rats that received a convulsive stimulation before each injection during the tolerance-dissipation phase, whether they were injected with saline or diazepam; in contrast, tolerance did not dissipate in the rats that received saline injections but no stimulations. Remarkably, the discontinuance of the bidaily diazepam injections, even for 16 days, was not sufficient to dissipate the tolerance that had developed to diazepam's anticonvulsant effect; nor was the continuation of the bidaily diazepam injections sufficient to keep tolerance from dissipating.(ABSTRACT TRUNCATED AT 250 WORDS)