The effects of denervation and long-term botulinum toxin type A (BoTx) poisoning on miniature end-plate potentials (m.e.p.p.s) in the frog were studied with intracellular microelectrode recording. BoTx reduced the frequency of m.e.p.p.s to less than 1% of the level seen in untreated frogs, leaving a large percentage of tiny m.e.p.p.s and slow-rising m.e.p.p.s (slow m.e.p.p.s). Unlike what is observed in the rat, the frequency of slow m.e.p.p.s never increased above the low rate measured in the untreated controls, and in fact slightly but significantly decreased after BoTx. A comparison of the m.e.p.p.s seen after BoTx poisoning (BoTx m.e.p.p.s) and m.e.p.p.s seen after denervation (Schwann m.e.p.p.s) revealed many similarities between the two including amplitude and time-to-peak distributions, temperature Q10 values and responses to several drugs and procedures. However, it was concluded that BoTx m.e.p.p.s do not originate from the Schwann cells because denervation of BoTx-paralysed frogs abolishes all m.e.p.p.s and the drug 4-aminoquinoline affects BoTx m.e.p.p.s and Schwann m.e.p.p.s in opposite ways, increasing the frequency of the former while almost eliminating the latter. BoTx m.e.p.p.s and Schwann m.e.p.p.s probably represent similar processes of secretion which are non-specific in nature, having a lower energy barrier than for normal release and not originating from specialized areas of transmitter release.