The usefulness of three retrograde fluorescent dyes for tracing injured peripheral axons was investigated. The rat sciatic was transected bilaterally and the proximal end briefly exposed to either Fast Blue (FB), Fluoro-Gold (FG) or to Diamidino Yellow (DY) on the right side, and to saline on the left side, respectively. The nerves were then resutured and allowed to regenerate. Electrophysiological tests 3 months later showed similar latencies and amplitudes of evoked muscle and nerve action potentials between tracer groups. The nerves were then cut distal to the original injury and exposed to a second (different) dye. Five days later, retrogradely labelled neurones were counted in the dorsal root ganglia (DRGs) and spinal cord ventral horn, The number of neurones labelled by the first tracer was similar for all three dyes in the DRG and ventral horn except for FG, which labelled fewer motoneurones. When used as second tracer, DY labelled fewer neurones than FG and FB in some experimental situations. The total number of neurotics labelled by the first and/or second tracer was reduced by about 30% compared with controls. The contributions of cell death as well as different optional tracer combinations for studies of nerve regeneration are discussed.