The steroidal module of the athlete biological passport (ABP) targets the use of pseudo-endogenous androgenous anabolic steroids in elite sport by monitoring urinary steroid profiles. Urine and blood samples were collected weekly during two consecutive oral contraceptive pill (OCP) cycles in 15 physically active women to investigate the low urinary steroid concentrations and putative confounding effect of OCP. In urine, testosterone (T) and epitestosterone (E) were below the limit of quantification of 1 ng/ml in 62% of the samples. Biomarkers' variability ranged between 31% and 41%, with a significantly lesser variability for ratios (except for T/E [41%]): 20% for androsterone/etiocholanolone (p < 0.001) and 25% for 5α-androstane-3α,17β-diol/5ß-androstane-3α,17β-diol (p < 0.001). In serum, markers' variability (testosterone: 24%, androstenedione: 23%, dihydrotestosterone: 19%, and T/A4: 16%) was significantly lower than in urine (p < 0.001). Urinary A/Etio increased by >18% after the first 2 weeks (p < 0.05) following withdrawal blood loss. In contrast, serum T (0.98 nmol/l during the first week) and T/A4 (0.34 the first week) decreased significantly by more than 25% and 17% (p < 0.05), respectively, in the following weeks. Our results outline steroidal variations during the OCP cycle, highlighting exogenous hormonal preparations as confounder for steroid concentrations in blood. Low steroid levels in urine samples have a clear negative impact on the subsequent interpretation of steroid profile of the ABP. With a greater analytical sensitivity and lesser variability for steroids in healthy active women, serum represents a complementary matrix to urine in the ABP steroidal module.