Correlational research suggests that men show greater attraction to feminine female faces when their testosterone (T) levels are high. Men's preferences for feminine faces also seem to vary as a function of relationship context (short versus long-term). However, the relationship between T and preferences for female facial femininity has yet to be tested experimentally. In the current paper, we report the results of two experiments examining the causal role of T in modulating preferences for facial femininity across both short and long-term mating contexts. Results of Experiment 1 (within-subject design, n = 24) showed that participants significantly preferred feminized versus masculinized versions of women's faces. Further, participants showed a stronger preference for feminine faces in the short versus the long-term context after they received T, but not after they received placebo. Post-hoc analyses suggested that this effect was driven by a lower preference for feminine faces in the long-term context when on T relative to placebo, and this effect was found exclusively for men who received placebo on the first day of testing, and T on the second day of testing (i.e., Order x Drug x Mating context interaction). In Experiment 2 (between-subject design, n = 93), men demonstrated a significant preference for feminized female faces in the short versus the long-term context after T, but not after placebo administration. Collectively, these findings provide the first causal evidence that T modulates men's preferences for facial femininity as a function of mating context.