BACKGROUND: The identification of low breastfeeding self-efficacy allows for tailored interventions to increase the rate of exclusive breastfeeding among women in accordance with WHO recommendations. While breastfeeding self-efficacy can be measured using the widely accepted Breastfeeding Self-Efficacy Scale-Short Form (BSES-SF), the construct and predictive validity of a translated German version have not been sufficiently demonstrated. The present study aimed to assess the psychometric properties of an alternative German translation of the BSES-SF. METHODS: The BSES-SF was administered to mothers at three months (N = 143) and six months (N = 115) postpartum. In addition, exclusive breastfeeding duration and related constructs were assessed. RESULTS: Excellent internal consistency (Cronbach's α = 0.91) was shown. A good model fit was achieved in the confirmatory factor analysis only when error covariances were considered (CFI = 0.95, SRMR = 0.057, RMSEA = 0.074), raising concerns about a unidimensional structure in the translated version. Construct and divergent validity were supported by significant correlations with self-esteem and social support, and negative correlations with depression, stress, and general anxiety. Importantly, predictive validity was confirmed. Furthermore, breastfeeding self-efficacy at three months postpartum was significantly correlated with exclusive breastfeeding duration at six months (r = 0.312). CONCLUSION: This study ensures the construct validity of the German BSES-SF and confirms its predictive validity for the first time. The German translation enables midwives and healthcare professionals in German-speaking countries to identify mothers in need of breastfeeding support up to three months postpartum. Targeted interventions can be tailored based on the overall BSES-SF score or specific item responses, contributing to improved breastfeeding outcomes.