The interspike interval with the lowest firing frequencies during repetitive firing to injected current in cat spinal motoneurones corresponds to the duration of the post-spike afterhyperpolarization (AHP) (Kernell, 1965). When recording from motor units in humans during slowly increasing and followed by decreasing contraction strengths it has been repeatedly confirmed that the lowest firing frequencies are seen at the end of the spike train (at de-recruitment with decreasing contraction) rather than just after recruitment (with increasing strength) (see Gorassini et al. 2002). Similar findings have been reported for motoneurones (intracellular recording) in the decerebrate cat when activated either with muscle stretch or intracellular current injections (Bennett et al. 1998). It has been suggested that the lower firing frequencies at de-recruitment may be explained by a sustained subthreshold depolarization with fluctuating synaptic noise occasionally driving the cell to fire (e.g. Matthews, 1996), rather than being determined by the AHP of the motoneurones as during a suprathreshold drive. To address this question experimentally we have recorded the AHP trajectory following spikes evoked by short intracellular current pulses. These test AHPs were recorded either alone (control) or following spike trains evoked by triangular current pulses (conditioned AHPs). The duration and amplitude of the triangular current pulses could be varied. The interval after the termination of the conditioning triangular current injection and the test pulse was varied between 250 ms and 2 s. Results were obtained from 10 cats. Initial surgery was performed under general anesthesia (Isofluran) and the cats were then decerebrated and neuromuscular blockade established with pancuronium (0.6 mg/h) (see Crone et al. 1988) in order to study the both primary and secondary range firing (the latter dependent on anaesthesia-dependent persistent inward currents). We confirmed that the firing frequency was lower at de-recruitment than at recruitment in all 51 neurones included in this report. We demonstrated that the AHPs were significantly prolonged in 30 motoneurones, when conditioned by a preceding spike train. This prolongation was often more pronounced at shorter than longer conditioning- test intervals (time between the end of the conditioning current injection to the test pulse) and with longer spike trains reaching high frequencies. We therefore conclude that the lower frequency at de-recruitment than at recruitment at least partly depends on a prolongation of the post-spike AHP.