IntroductionTestosterone administration attenuates reductions in total body mass and lean mass during severe energy deficit (SED).ObjectivesThis study examined the effects of testosterone administration on the serum metabolome during SED.MethodsIn a double-blind, placebo-controlled clinical trial, non-obese men were randomized to receive 200-mg testosterone enanthate/wk (TEST) (n = 24) or placebo (PLA) (n = 26) during a 28-d inpatient, severe exercise- and diet-induced energy deficit. This study consisted of three consecutive phases. Participants were free-living and provided a eucaloric diet for 14-d during Phase 1. During Phase 2, participants were admitted to an inpatient unit, randomized to receive testosterone or placebo, and underwent SED for 28-d. During Phase 3, participants returned to their pre-study diet and physical activity habits. Untargeted metabolite profiling was conducted on serum samples collected during each phase. Body composition was measured using dual-energy X-ray absorptiometry after 11-d of Phase 1 and after 25-d of Phase 2 to determine changes in fat and lean mass.ResultsTEST had higher (Benjamini–Hochberg adjusted, q < 0.05) androgenic steroid and acylcarnitine, and lower (q < 0.05) amino acid metabolites after SED compared to PLA. Metabolomic differences were reversed by Phase 3. Changes in lean mass were associated (Bonferroni-adjusted, p < 0.05) with changes in androgenic steroid metabolites (r = 0.42–0.70), acylcarnitines (r = 0.37–0.44), and amino acid metabolites (r = − 0.36–− 0.37). Changes in fat mass were associated (p < 0.05) with changes in acylcarnitines (r = − 0.46–− 0.49) and changes in urea cycle metabolites (r = 0.60–0.62).ConclusionTestosterone administration altered androgenic steroid, acylcarnitine, and amino acid metabolites, which were associated with changes in body composition during SED.