While breast self-examination (BSE) frequency has received extensive research attention, proficiency has been less frequently addressed. Moreover, BSE proficiency among women at increased risk has not been adequately examined. Assessment of BSE proficiency is critical in determining the value of BSE with mammography and clinical breast examination in the early detection of breast cancer. BSE proficiency was assessed in 101 first-degree relatives of breast cancer patients. Participants were stratified by BSE frequency and randomized to one of two training techniques (MammaCare or concentric circle). BSE performance was assessed at baseline and at three follow-up visits at 4-month intervals. Proficiency was assessed by verbal description, a projected grid observational method, and lump detection ability on two breast models. BSE frequency was also assessed, in addition to BSE confidence, knowledge of breast cancer, risk perception, and worry related to breast cancer development. At baseline, proficiency was poor and correlations were not significant across assessment modalities. Significant improvement occurred on self-report measures, lump detection ability (true positives) on both models, and the projected grid. Improvement occurred across both training groups by the first follow-up, with no changes at subsequent visits. Both training techniques significantly improved BSE proficiency and were viewed positively by participants.