The abilities of the fluorescent retrograde tracers true blue and diamidino yellow to label motor neurons of the rat sciatic nerve were compared quantitatively. Following injection of a mixture of the 2 tracers into the sciatic nerve, diamidino yellow was found only in double-labelled neurons, while 28% of labelled neurons contained true blue alone. The relative labelling efficiency of diamidino yellow, at only 72%, was significantly lower than that of true blue. When the tracers were injected separately a difference in the labelling efficiency was still observed but, in addition, there were significantly fewer diamidino yellow-labelled neurons than when a mixture had been injected. This suggests that the presence of true blue in the mixture had enhanced the uptake, transport or visualisation of diamidino yellow. When a mixture of true blue and diamidino yellow was applied to the cut sciatic nerve, the relative labelling efficiency of diamidino yellow (77%) was again found to be lower than that of true blue, but positive identification of diamidino yellow-labelled cells was hampered by chromatolytic changes in the cell bodies. Injection of the tracer mixture into the gastrocnemius muscle resulted in a diamidino yellow labelling efficiency (36%) significantly lower than that obtained with either nerve injection or nerve dipping. Thus, compared to true blue, diamidino yellow was either less capable of reaching the motor endplates within the muscle, or it was taken up less efficiently by axon terminals than by the axons themselves.