Optimal selective stimulation of nerve with endoneurally (intrafascicularly) inserted multi-microelectrodes means that each electrode activates, with its own threshold stimulation current, as few distinct motoneurons as possible, preferibly only one. If the latter is the case, the efficiency of a multi electrode is 100%. However, neighbouring electrodes may control the same motor fiber(s), as there are generally more fibers than electrodes and because the position of fibers is largely unknown. In that case, efficiency is less than 100%. This paper reports on experiments in rat peroneal nerve with 5-and %-fold wire multi micro electrode arrays. The threshold force of the twitch recruitment curve of the corresponding EDL muscle was used to monitor nerve activation. It was found that on average the threshold force efficiency was 0.48 = 48 %. After re-inspection of the data, taking into account that neighbouring electrodes have a higher probability to activate the same motor units, in contrast to distant electrodes, the average efficiency even rises to 81 %. For several reasons, threshold forces do not correspond to motor unit forces, implying that the threshold-forceefficiency can not be regarded as motor-unit-efficiency.