Background: Women treated for rectal cancer are at risk of sexual dysfunction and impaired ovarian androgen production. Aim: To investigate a possible association between serum levels of endogenous androgens and sexual function in women with rectal cancer. Methods: Women diagnosed with stage I−III rectal cancer were consecutively included and prospectively followed with the Female Sexual Function Index (FSFI) questionnaire from baseline to 2 years postoperatively and blood samples for hormone analyses, baseline to 1 year. Androgens were measured with liquid chromatogra-phyÀmass spectrometry and electrochemiluminescence. The associations between the 4 measured androgens (testosterone, free testosterone, androstenedione, and dehydroepiandrosterone sulphate) and sexual function were assessed with generalized least squares random effects regression analysis in sexually active women. Outcomes: The primary outcome measure was the mean change observed in the FSFI total score when the serum androgen levels changed with one unit. Secondary outcomes were the corresponding mean changes in the FSFI domain scores: sexual desire, arousal, lubrication, orgasm, satisfaction, and pain/discomfort. Results: In the 99 participants, the median FSFI total score decreased from 21.9 (range 2.0 − 36.0) to 16.4 (3.5 − 34.5) and 11.5 (2.0 to 34.8) at 1 and 2-years follow-up. After adjustment for age, partner, psychological well-being, preoperative (chemo)radiotherapy, and surgery, total testosterone and androstenedione were significantly associated with FSFI total score (b-coefficients 3.45 (95% CI 0.92 − 5.97) and 1.39 (0.46 − 2.33) respectively). Testosterone was significantly associated with the FSFI-domains lubrication and orgasm, free testosterone with lubrication, androstenedione with all domains except desire and satisfaction, and dehydroepiandrosterone sulphate with none of the domains. Strengths and Limitations: This is the first study investigating whether androgen levels are of importance for the impaired sexual function seen in women following rectal cancer treatment. The prospective design allows for repeated measures and the use of the FSFI for comparisons across studies. No laboratory data were collected at the 2-year follow-up, and the missing data could have further clarified the studied associations. Conclusion and Clinical Implication: Testosterone and androstenedione were associated with sexual function in female rectal cancer patients. The results are of interest for future intervention studies and contribute to the understanding of sexual problems, which is an essential component of the rehabilitation process in pelvic cancer survivors.