Fifty 2- to 16- yr old pony stallions were randomly assigned to one of five treatments: Group 1, controls (no treatment); Group 2, mg/kg stanozolol weekly for 13 treatments; Group 3, mg/kg stanozolol every 3 wk for 5 treatments; Group 4, mg/kg boldenone undecylenate every 3 wk for 5 treatments; and Group 5, boldenone undecylenate weekly for 13 treatments. Mean plasma testosterone levels for Groups 2, 4, and 5 were elevated over controls (P<) at 2, 8, and 9 wk, respectively. Testosterone levels for ponies in Group 3 did not differ from controls (P>). There were no differences in mean plasma luteinizing hormone (LH) and follicle stimulating hormone (FSH) levels among groups (P>). Daily spermatid production per gram of testicular parenchyma (DSP/gm) in Group 5 was lower than in controls (P<), whereas DSP/gm was not different among groups 1 to 4 (P>). There were no differences among groups (P>) in the percentage of Stage 8 tubules or relative number of Leydig cells. The mean diameter of Leydig cells was less for Group 5 than for controls (P<), but was not different for Groups 1 to 3 (P>).