While in vitro studies on the transgenic SOD1-G93A mouse model of human amyotrophic lateral sclerosis (ALS) provided valuable insights into the molecular and cellular disruptions underlying familial ALS, little is known about the significance of those observations for the in vivo situation of the intact animal. To elucidate underlying mechanisms we investigated lumbar spinal reflexes in SOD1-G93A mice in vivo and compared the results to those from wild type (WT) littermates. Adult (>90 days) WT mice of strain B6SJL and transgenic SOD1-G93A mice (~90 days) were used: initial anaesthesia with pentobarbital sodium 70 mg kg-1 i.p., continuation with i.v. infusion of methohexital sodium 40-60 mg kg-1 h-1 as required. Body core temperature, ECG and breathing volume and rate were controlled throughout the experiment. A laminectomy was performed from L1 to L5. Nerves of the left hind limb (posterior biceps, common peroneal, tibial and sural nerve) were prepared, distally cut, and mounted for recording of monosynaptic reflexes. The left dorsal root L4 (main input from the hind leg) was cut and mounted for central stimulation (stimuli: rectangular pulses of 0.1 ms duration, strength 2-10 times threshold). The recurrence interval of the stimulus started with 1.84 s, and was gradually reduced to 50 ms. The reduction of monosynaptic reflexes with increasing stimulus frequency and the required recovery time of the reflex after repetitive stimulation were determined. Before the experiments there were different degrees of motor dysfunctions in the SOD1-G93A mice. WT mice showed a stable reflex amplitude with 1.84 s stimulus interval, and a decreasing reflex amplitude during higher stimulus frequencies. In all animals, recovery of reflexes was complete within a few minutes. SOD1-G93A mice with light motor deficits showed a distinct reduction of the monosynaptic reflex amplitude during repetitive stimulation, even with long stimulus intervals. In mice with severe motor deficits, a monosynaptic response could hardly be elicited. In general, recovery of monosynaptic reflexes after repetitive stimulation required a longer time in SOD1-G93A mice and was mostly incomplete. In SOD1-G93A mice without obvious motor deficits, the monosynaptic reflex amplitude was normal with 1.84 s stimulus interval. Although recovery times were longer compared to WT, recovery was in most cases complete. The experiments demonstrate a clear deficit in the monosynaptic reflex of SOD1-G93A mice, reflex profiles correlating with apparent motor deficits. The results on reflex recovery i) are in agreement with models suggesting a contribution of impaired cellular energy supply to the pathogenesis of ALS and ii) provide a basis for future in vivo studies on ALS-related spinal disturbances.