Long-Latency-Reflexes (LLR) are indirect, sensorymotor evoked potentials. Methods: LLR were registered over the slightly activated M. biceps brachii (BB) und M. abductor digiti minimi (ADM) on either side simultaneously with unilateral supramaximal 1/s-stimuli over the N. radialis superficialis at the wrist (8 males, 20 to 42 yr.). 60 responses were averaged and reproduced once; 400 ms time window. Trains of 3 (3S) und 5 stimuli (5S) (inter-stimulus intervals: 3 ms) were compared with single stimuli (1S). Results: The ipsilateral BB showed at least 2 reproducible responses M1 (n = 7) und M2 (n = 8). Further 1 to 3 responses were facultatively observed (n = 3). Amplitudes were: M1, 1S, 78 ± 47 μV (mean ± SD); 3S, 122 ± 46 μV (p < 0,005); 5S, 155 ± 91 μV (p < 0,05). M2, 72 ± 56 μV (1S); 134 ± 85 μV (3S; p < 0,01), 197 ± 112 μV (5S; p < 0,01). Onset-latencies were not significantly different (mean M1, 50 ms; M2, 85 ms). Over the ipsilateral ADM M1 and M2 were seen in 3 (1S) or 7 cases (5S). Over the contralateral muscles responses were less frequent (n = 2 for the BB, n = 0 for the ADM; 1S), (n = 4 for the BB, n = 3 for the ADM; 5S), in part only M1, in part M1 and M2. Discussion: Amplification of LLR by trains of stimuli infers increasing numbers of recruited supraspinal motoneurones by subsequent stimuli. Responses contralateral to the stimulus may be due to the known ipsilateral projections of the motor cortex. The repetitive discharges (usually 2, in single cases up to 5) may reflect characteristics of central motor neurons, that have not yet been seen after direct stimulation over the motor cortex. The frequency of the subsequent responses lies within the beta-band. At present we investigate whether the frequency of LLR-discharges is coherent with somatosensory evoked potential responses or with beta-frequencies in EEG power spectral analysis.