Introduction: Correctly identifying pathology in pre-pubertal females is a high-stakes physical examination skill. Currently, learning this skill relies heavily on case-by-case exposure, which is variable, limited and often results in suboptimal skill. Thus, there is a need to develop and evaluate learning platforms that simulate the presentation and diagnosis of this important clinical task. We developed an on-line learning and assessment platform that allowed the deliberate practice of 158 pre-pubertal female genital image interpretations . We examined the quantity of skill acquisition by deriving performance metrics and learning curves. Methods: This was a prospective cross-sectional study administered via an on-line learning and assessment platform. Colposcopic images were acquired from a child abuse clinic. Two child abuse experts interpreted images to determine case solutions and 40% of cases had medical or traumatic pathology. Further, to validate image interpretations, a unique set of five child abuse and pediatric gynaecology experts reviewed the cases. Study participants were recruited from the USA and Canada and were required to complete all 158 cases. For each image, learners designated cases as normal or abnormal and if abnormal indicated the abnormal area on the image. The primary outcome was the change in accuracy, sensitivity and specificity. Results: We enrolled 107 participants, 26 medical students, 31 pediatric residents, 24 pediatric emergency fellows, and 26 pediatric emergency attendings. For all participants, the change in accuracy was +9.6% for accuracy (<0.001), +1.4% for sensitivity (p=0.6) and +15.7% (p<0.001) for specificity. The final score for accuracy, sensitivity and specificity was 79.5%, 66.1%, and 87.8%, respectively. There was no difference between learner types with respect to summary performance metrics (accuracy, p=0.15; sensitivity, p=0.44; specificity, p=0.54). Learning curves show maximal learning gains (inflection point) up until 100 cases. Conclusion: Deliberate practice of pre-pubertal female image interpretation was effective for ensuring predictable skill improvement for normal cases but was less effective for abnormal cases. Future research could examine how to refine the education tool to better serve diagnostic skill of abnormal cases.