The quantitative matching hypothesis of neuronal cell death was tested for the chick hindlimb by determining the relationship between myotube number at the onset of motoneuron cell death and the number of motoneurons that survive in chicks, quail, and chick-quail chimeras. Hindlimb buds, which differ in size between the 2 species, were exchanged at stages 16 1/2-19, myosin ATPase-stained myotubes in selected thigh muscles were counted during the cell death period (stages 30-34), and lumbosacral motoneurons were counted following the cell death period (stage 38). No quail motoneurons were rescued when quail cords innervated chick limbs. When chick cords innervated quail limbs, the number of surviving motoneurons was significantly decreased but not to quail values. We consider that this occurred because chicks develop more slowly than quail, and we found that transplanted chick limbs were developmentally younger than the contralateral quail limb at the onset of motoneuron cell death and contained fewer myotubes. Similarly, transplanted quail limbs contained more myotubes at the onset of cell death than normal stage 30 quail limbs. An excellent correlation was obtained during normal development of both species between the number of myotube clusters at the onset of cell death and the number of surviving motoneurons. This correlation was also observed for chick-quail chimeras, and when the data points were plotted for control chick, control quail, chick host-quail limb, and quail host-chick limb, the correlation coefficient was 0.996. This strongly suggests that some parameter closely related to myotube number limits the number of motoneurons that will survive. A proposal consistent with our observations is that motoneuron survival is dependent on the uptake of a myotube-derived trophic factor that can only be taken up at synaptic sites and that the number of such sites is limited and directly related to myotube number. In conclusion, our observations strongly support a quantitative-matching component in the process of neuronal cell death. However, since we were unable to rescue any neurons, we cannot exclude the possibility that some proportion of neurons normally dies for reasons other than peripheral competition.