Background: Infants who are not exclusively breastfed during the first six months of life are given a wide variety of other foods or drinks during the period to complement or replace breast milk, these food items are termed Breast Milk Substitutes (BMS). Inadequate Exclusive Breastfeeding (EBF) is associated with increased infant morbidity and death. In Ghana about one-third of infants under six months old are not exclusively breastfed suggesting reliance on breast milk substitutes (BMS) for infant feeding. Factors associated with mothers? use of BMS has not been adequately explored in Ghana to inform effective interventions to improve EBF and duration. Objective: To assess the prevalence, determinants and timing of use of BMS among mothers of children 3 to 6 months attending child welfare clinics (CWC) in 3 selected health facilities in the Ledzokuku –Krowor Municipality. Methodology: A cross-sectional survey with 360 mother infant pairs recruited from CWC at three hospitals (Mission, Public, Private). Data on socio-demographic characteristics, Exclusive Breastfeeding Self Efficacy (EBFSE), exposure to promotion of BMS and use of BMS were collected via face-to-face interviews using semi structured questionnaires. Bivariate analyses were used to assess significant differences between mothers who used and those who did not use BMS in infant feeding. Binary logistic regression analysis was used to identify factors that predicted mothers? use of BMS. Results: The prevalence of use of BMS was 51.9% and the mean duration of EBF was 2.1 ±1.6 mos among infants who had been introduced to BMS. The main BMS used were water and/or gripe, infant formula and complementary foods. Mothers? main source of exposure to the promotion of BMS was through advice (50.6%); with health workers being the primary source of advice (74.2%) on using BMS. Significantly more mothers who used BMS than those who did not had been exposed to the promotion of BMS (67.9% vs 55.5%, p = 0.015) through commercial advertising or advice from others. Mothers who had already introduced BMS to their infants had a lower mean EBFSE score than those who were not using BMS (20.54 ±10.52 vs. 33.52±4.53), p = 0.001) suggesting that caregivers? confidence in practicing EBF influenced the use of BMS. Mothers who did not work away from their infants were less likely to use BMS than those who worked away from their infants (OR 0.142; CI, 0.047 - 0.429); mothers who planned to exclusively breastfeed before delivery were less likely to use BMS than those who planned to mix feed (OR – 0.137; CI, 0.031 - 0.613) while mothers who had a lower EBFSE score had 40.6 times higher odds (95%CI= 15.760 to 104.466) of using BMS. Conclusion: Infant feeding interventions in the Ledzokuku-Krowor Municipality should be directed at improving maternal confidence in practicing EBF since women with lower confidence are more likely to use BMS. These is also a need for significant effort to change health worker attitude towards the use of BMS among infants less than 6 months.