Single muscle fibre action potentials (SFAPs) are considered to be functions of a bioelectrical source and electrical conductivity parameters of the medium. In most model studies SFAPs are computed as a convolution of the bioelectrical source with a transfer function. Calculated peak-to-peak amplitudes of SFAPs decrease with increasing recording distances. In this paper an experimental validation of model results is presented. Experiments were carried out on the m. extensor digitorum longus (EDL) of the rat. Using a method including fluorescent labelling of the active fibre, the distance between the active fibre and the recording electrode was derived. With another method, the decline of the peak-to-peak amplitude of SFAPs detected along a multi-electrode was obtained. With both experimental methods, in general peak-to-peak amplitudes of SFAPs decreased with increasing recording distances, as was found in model results with present volume conduction theory. However, this behaviour was not found in all experiments. The rate of decline of the peak-to-peak amplitudes with recording distance was always less than in models.