Digit ratio (2D:4D) is widely used as a marker of prenatal androgen exposure. However, there are no published prospective studies where prenatal androgen exposure has been measured and correlated with digit ratio in adult life. We aimed to establish the prospective relationship between prenatal androgen exposure in the second and third trimesters of pregnancy (as measured by maternal circulating androgen concentrations and umbilical cord androgen concentrations) and digit ratio in adolescent girls. Androgen concentrations (testosterone, free androgen index, androstenedione, DHEAS) and sex hormone binding globulin (SHBG) were measured in stored plasma samples from pregnant mothers at 18 (n=118) and 34/36 (n=114) weeks of gestation and in cord blood (n=82) from the Western Australian Pregnancy (Raine) Cohort Study (www.rainestudy.org.au). Digit ratio was measured in 244 female offspring from this cohort at age 14-16 years. Only one borderline statistically significant correlation between maternal circulating androstenedione levels at 18 weeks of gestation and left hand digit ratio was seen. No other statistically significant relationship between maternal androgen concentrations or umbilical cord androgen concentrations and digit ratio in adolescence were observed. These findings suggest that variation in 2D:4D in girls is not established as a result of testosterone concentrations in the second and third trimesters. We conclude that prenatal androgen exposure as measured by maternal circulating androgen concentrations at 18 and 34/36 weeks of gestation or in the umbilical cord at birth may not predict digit ratio in girls.