Motoneurons seem to require contact with their target muscle even after embryogenesis is complete, but the consequences of target-deprivation during postnatal development are poorly understood. To examine the fate of motoneurons separated from their targets postnatally, we labeled the motoneurons that innervate the biceps brachii muscle with the retrograde tracer Fluorogold and then separated them from their muscle by amputating the forelimb. Fluorogold was subsequently found within motoneurons, as well as within much smaller cells that were identified as microglia. The number of labeled microglial cells steadily increased with time following limb amputation, while the number of labeled motoneurons declined. The magnitude of this response depended on the age of the animal: the younger the animal at the time of the amputation, the greater the number of labeled microglia and the more extensive the neuronal loss. To ensure that the response to amputation was caused by target deprivation, rather than by the injury itself, the nerve to the biceps muscle was cut or crushed. In this way, axons were transected but target access was only temporarily denied. After the nerve was cut, motoneurons began to reinnervate the muscle within 3 weeks but, just as after amputation, the spinal cord subsequently contained labeled microglia and a reduced number of motoneurons. In contrast, after nerve crush, reinnervation began within 4 d and there was no evidence of motoneuron death. Our results demonstrate that target-deprivation causes motoneurons to be lost in an age- and time- dependent manner, and indicate a critical period after axotomy during which motoneurons must reinnervate their target in order to survive. Further, we provide evidence that microglial cells may phagocytose dying motoneurons. The approach we used would provide a convenient assay for testing candidate motoneuron growth factors in animals where in vivo studies of the embryo are difficult.