The objective of this study was to determine whether decreased gonadal function is a risk factor for hip fracture in elderly men. The study was a matched case-control study performed at an in-hospital orthopedic service at a semi-rural and academic tertiary care center. The patients were seventeen men who presented with hip fractures after simple falls over a 10-month period, 11 men with a history of hip fractures in the preceding 25 months, and 28 randomly selected age-, race-, and living status-matched control subjects (mean age 73 years). Serum pooled total testosterone (9.2 +/- 5.5 nmol/L vs. 12.8 +/- 5.4 nmol/L; p less than 0.02) and free testosterone (37.9 +/- 18.8 pmol/L vs. 48.4 +/- 20.6 pmol/L; p less than 0.03) were significantly lower in hip fracture vs. control patients. Testosterone deficiency was found in 20 (71%) hip fracture men vs. 9 (32%) of the controls (p = 0.003; odds ratio 5.3). Analysis of testosterone values within the stratum of absence of any chronic disorder revealed similar results. Serum 25-hydroxyvitamin D levels were significantly lower in the hip fracture group than in control men (p less than 0.001). The conclusion is that gonadal deficiency appears to be an important and heretofore understudied risk factor for hip fractures in men. Prevention of hip fractures in men may involve early recognition and treatment of testosterone deficiency.