This prospective study evaluated the 3-year change in menstrual function and bone mass among 40 female adolescent endurance runners (age 15.9 ± 1.0 years) according to baseline disordered eating status. Three years after initial data collection, runners underwent follow-up measures including the Eating Disorder Examination Questionnaire and a survey evaluating menstrual function, running training, injury history, and prior sports participation. Dual-energy X-ray absorptiometry was used to measure bone mineral density and body composition. Runners with a weight concern, shape concern, or global score ≥4.0 or reporting >1 pathologic behavior in the past 28 days were classified with disordered eating. Compared with runners with normal Eating Disorder Examination Questionnaire scores at baseline, runners with disordered eating at baseline reported fewer menstrual cycles/year (6.4 ± 4.5 vs. 10.5 ± 2.8, p = .005), more years of amenorrhea (1.6 ± 1.4 vs. 0.3 ± 0.5, p = .03), and a higher proportion of menstrual irregularity (75.0% vs. 31.3%, p = .02) and failed to increase lumbar spine or total hip bone mineral density at the 3-year follow-up. In a multivariate model including body mass index and menstrual cycles in the past year at baseline, baseline shape concern score (B = -0.57, p value = .001) was inversely related to the annual number of menstrual cycles between assessments. Weight concern score (B = -0.40, p value = .005) was inversely associated with lumbar spine bone mineral density Z-score change between assessments according to a multivariate model adjusting for age and body mass index. These finding support associations between disordered eating at baseline and future menstrual irregularities or reduced accrual of lumbar spine bone mass in female adolescent endurance runners.