Of eight groups of five adult male Sprague-Dawley rats, three groups received an intraperitoneal dose of EDS (75 mg kg-1) in DMSO (dimethylsulfoxide in water; 1:3) and three control groups received DMSO only. One EDS and one control group were killed 1, 2 and 3 days after treatment. One EDS + T group, given a 5-cm implant of T sufficient to maintain normal serum T concentrations, and one untreated control group were killed on Day 3. Epididymides were fixed by vascular perfusion and embedded in Epon/Araldite for light and electron microscopic studies. Epididymal duct diameters and epithelial heights were measured from l-/im sections and structural changes were assessed from thin sections using a Jeol 100B electron microscope. Morphological studies showed a reduction in duct diameter and an associated increase in epithelial height in the proximal cauda epididymidis of the experimental groups. In the EDS + T implant group, epithelial heights were significantly greater than in controls but duct diameters remained unchanged. The most obvious structural change in the proximal cauda epididymidis was the selective destruction of clear cells in the epithelium. Initially, vacuoles were observed in the lateral intercellular spaces of the epithelium; large autophagic vacuoles then appeared in the clear cells, which had degenerated and disappeared from the epithelium by Day 3. Progressive infiltration of leucocytes into the intertubular interstitium, the epithelium and lumen of the proximal cauda was also observed. Loss of clear cells in the proximal cauda epididymidis was also seen in the EDS + T group, suggesting that clear-cell degeneration was not associated with reduced concentrations of circulating androgen. In all EDS-treated groups, however, clear cells and duct profiles in the distal cauda epididymidis remained unaffected. The reason for their protection from the effects of EDS has not yet been determined. These results suggest that, in addition to other specific lesions described in previous studies, EDS also has a direct effect on the rat epididymis that appears to be specifically targeted to the clear cells in the proximal caudal region.