OBJECTIVE: To report range of motion at the hip and ankle in male and female ballet dancers compared with controls. DESIGN: Cross-sectional cohort study with convenience control sample. SETTING: National classical ballet school in Australia. PARTICIPANTS: Thirty-three female and 30 male full-time ballet students of mean (SD) age 16.9 (0.8) and 18.0 (1.4) years, respectively. Controls were 31 female and 16 male nondancing, normally active university students of average ages 18.8 +/- 0.6 and 18.8 +/- 0.9 years, respectively. MAIN OUTCOME MEASURES: Degrees of range of motion of left and right sides for the following four movements: standing plié in parallel-passive ankle dorsiflexion (DF), standing turnout in the balletic first position-lower limb external rotation (LLER), supine hip external rotation (ER), supine hip internal rotation (IR). Two additional ranges of motion were calculated. External rotation below the hip joint (BHER) was derived by subtracting hip ER from LLER, and total supine hip rotation (Hip IR + ER) was derived by summing hip ER and hip IR. MAIN RESULTS: On both left and right sides, dancers had greater LLER (p < 0.0001) and hip ER (p < 0.0001) but lesser hip IR than the controls (p < 0.0001). Female but not male dancers had greater total supine hip rotation (hip ER+IR) than controls (p < 0.05). There was no difference in either BHER or ankle DF between dancers and controls. Within dancers, females had greater range in both left and right hip IR, LLER, and hip IR + ER. There was no difference in hip ER. Regression analysis revealed that LLER related to hip ER in both males and females. CONCLUSIONS: Elite classical dance students had greater hip external rotation (both ER and LLER) and less hip internal rotation than controls but similar BHER and ankle DF. Since ER below the hip joint and ankle DF do not differ between dancers and controls, despite repeated training of these movements in ballet, there may be anatomical (bony/ligamentous) limitations to these movements.