Immunocytochemistry was used to localize the populations of tyrosine-hydroxylase-like (TH)-immunoreactive cells in the tiger salamander retina. Ninety percent of these cells possessed somas that were situated in the innermost cell row of the inner nuclear layer and were classified as amacrine cells. Ten percent of TH-immunoreactive somas were located in the ganglion cell layer and were tentatively designated as those of displaced amacrine cells. The processes of TH-immunoreactive cells ramified most heavily in sublayer 1 of the inner plexiform layer, while a relatively small number of TH-labelled processes distributed in sublayers 3 and 5. Less than 1% of TH-immunoreactive cells in the amacrine cell layer exhibited a short process of somal origin that extended distally toward the outer plexiform layer. However, these processes did not cross the whole of the inner nuclear layer, and no immunolabelling was observed in the outer plexiform layer. An examination of retinal whole-mounts revealed that TH-immunoreactive amacrine and displaced amacrine cells were distributed throughout the center and periphery of the retina. The density of TH-immunolabelled amacrine cells was calculated to be 49 +/- 13 (mean +/- standard error) cells per mm2. The vast majority of TH-immunoreactive amacrine and displaced amacrine cells exhibited a stellate appearance and gave rise to three or more primary dendrites. A few TH-amacrine and displaced amacrine cells possessed two primary dendrites that emerged from opposite sides of their somas. The processes of TH-immunoreactive cells were generally poorly branched and varicose with terminal branches sometimes appearing thin and beaded. Because some TH-immunolabelled processes were very long, there was considerable overlap between the dendritic fields of neighboring TH-cells. Lastly, individual TH-immunoreactive amacrine and displaced amacrine cells were often observed in whole-mounts to provide processes that ramified at more than one level of the inner plexiform layer.