We studied 14 postpubertal patients at an average of 33 months after treatment for testicular torsion. Of these patients 11 had been treated by detorsion and 3 by orchiectomy. Five normal male volunteers of the approximate age of the study group served as controls. The patients treated by detorsion were subdivided into 3 groups based on the degree of atrophy of the detorsed testicle: group 1--no testicular atrophy (5), group 2--25 per cent testicular atrophy (2) and group 3--greater than 90 per cent testicular atrophy (4). Mean duration of torsion was greatest in the orchiectomy group (161 hours) compared to 6, 16 and 29 hours for groups 1, 2 and 3, respectively. The serum luteinizing hormone and follicle-stimulating hormone response to an intravenous bolus of 100 mcg. synthetic gonadotropin releasing hormone was measured in all patients. All groups had a greater mean follicle-stimulating hormone response to gonadotropin releasing hormone stimulation than controls (p less than 0.05). Patients who underwent orchiectomy had the greatest follicle-stimulating hormone response to gonadotropin releasing hormone stimulation. Mean luteinizing hormone response to gonadotropin releasing hormone stimulation was normal in patients without atrophy (group 1) but it was greater than controls in patients who had atrophy (groups 2 and 3) or who underwent orchiectomy (p less than 0.05). Several conclusions could be made from our study. All patient groups treated for torsion had evidence of testicular dysfunction. Patients who underwent orchiectomy displayed more testicular dysfunction than patients who had atrophy after detorsion. Testicular dysfunction after torsion is more likely to involve spermatogenic before Leydig cell function.