Purpose: This study aimed to elucidate the prevalence of breast feeding and maternal psychosocial predictors affecting the breast feeding practice at six months postpartum. Method: One hundred and sixty one mothers at six months postpartum were conveniently recruited in two public health center in Jeju-city. Subjects completed a well-structured questionnaire regarding the feeding types of the child, maternal self-esteem, the Edinburgh Postnatal Depression Scale (EPDS), husband support, and marital adjustment scale. The data was analyzed using the $X^2-test$, t-test, Pearson correlation coefficients, and logistic regression. Result: The prevalence of breast feeding at six months postpartum was 29.9%. Predictors of breast feeding practice identified by the logistic regression analysis include first feeding type after delivery, marital adjustment, readiness for maternal role and work state. Conclusion: The prevalence of breast feeding practice was lower than WHO desired at six months postpartum. Therefore, the findings support the need for breast feeding education including interventions focused on improving readiness for maternal role and marital adjustment.