Background Electromyography (EMG) may give important information in addition to clinical findings in various neuromuscular disorders. Beside spontaneous discharges, motor unit potential analysis is routinely used to assess the size of motor units. Normative values for motor unit potentials (MUP) have been reported for e.g. biceps brachii or vastus muscles. Hardly any information is, however, available about motor unit potential (MUP) parameters of paraspinal muscles. Such information might be of special value, since in patients with myopathies or with spinal muscular atrophy (SMA), a proximal pattern of muscular involvement is the rule. Therefore, the aim of the present investigation was to obtain normative values for MUP parameters of the erector trunci (ET) muscle. Methods Thirtytwo healthy volunteers gave informed consent (O 28.9 ± 9.9 years; 16 women) to investigate the left biceps brachii (BB), vastus lateralis (VL) and ET muscles. BB and VL were investigated for comparative purposes. All EMG investigations and MUP analyses were done by conventional technique by MEK. For the BB and VL, subjects rested in the supine position. For recording the ET, patients were positioned on the right side with their hips, trunk and head in a maximally bent position. The needle electrode was inserted between lumbar vertebrae 1 and 2, two fingers lateral to the midline at a depth of 2.5–4.5 cm. MUP were recruited by extending head and legs simultaneously. MUP data comprized: amplitude (peak-to-peak), duration, area, phase number and size-index. Statistics: Kruskall–Wallis test, Mann-Whitney-U test post hoc . Results A total of 1683 MUP from 96 muscles were available. No statistically significant differences were noted for MUP parameters were detected between females and males. Therefore, values were treated together. Statistical analyses revealed significantly higher values in the VL for amplitude (0.56 ± 0.13 mV), duration (11.9 ± 1.0 ms), area (0.86 ± 0.22 μVs) and size-index (0.94 ± 0.24) in relation to BB: amplitude 0.37 ± 0.05 mV, duration 10.2 ± 0.9 ms, area 0.54 ± 0.11 μVs, size-index 0.52 ± 0.24, and ET: amplitude 0.39 ± 0.07 mV; duration 10.6 ± 1.2 ms; area 0.59 ± 0.16 μVs; size-index 0.60 ± 0.25. For these parameters no significant differences were observed between BB and ET. However, phase numbers were significantly smaller in BB (2.96) as compared with VL (3.18) and ET (3.22). Discussion MUP values in BB and VL are well in line with previously reported values. Therefore, the MUP data given for the ET here may be regarded as valid, too. ET MUP display similar values like those of the BB. Significantly higher values have been measured in the VL, with the exception of phase numbers. Amplitude, duration, area and size-index all depend on the amount of muscle fibres of a motor unit in the close vicinity of the needle electrode. Therefore, it can be inferred that ET motor units have similar dimension like those of the BB, while VL motor units appear to be significantly larger. The smaller dimensions of BB motor units relative to those of the VL may be clear from differences in their functions. For ET, which serves simple and mainly static movements of the vertebral column, larger dimensions of the motor units might have been expected than actually found here. Our results on ET MUP parameters are important for routine clinical practice because they facilitate the diagnosis in patients with suspected neuromuscular disorders such as myopathy or SMA.