Objective To determine (1) the frequency of testosterone (T) use in 2 academic male infertility practices, (2) the effect on semen parameters, and (3) the potential for reversibility. Methods We performed a retrospective review. Inclusion criteria were supplemental T usage at the time of initial visit, T cessation, and semen analyses before and after T cessation. Results From January 2005 to March 2011, 110 of 1540 patients evaluated for infertility (7.1%) were on supplemental T. Thirty-nine patients met the inclusion criteria. Twenty-two of 39 patients (56%) were started on medical recovery by medical treatment. Median sperm concentration increased from 0 to 6.3 million/mL (P <.0001) after T cessation, with median follow-up of 4.5 months. Eight men (21%) remained azoospermic, and 3 men had sperm concentrations of ≤0.1 million/mL. Fifteen of 22 men given medical treatment and 16 of 17 men given no treatment had sperm return to the semen, respectively (P =.1061). Final sperm concentration did not differ between those who received medical treatment and those who did not. Conclusion Exogenous T treatment causes infertility and is common among reproductive-aged men. These effects are transient for some, but not all men had return of sperm to the semen during the study interval, suggesting a possible underlying primary testicular dysfunction or ongoing hypothalamic-pituitary-testicular axis suppression. Education about exogenous T's negative reproductive effects is necessary to avoid this preventable form of infertility. Men with low T levels should have sufficient evaluation to evaluate for underlying pathology before starting T.