Background: Evidence of interventions that are effective in improving exclusive breastfeeding (EBF) practices is needed to help countries revise their strategies. To assess whether mothers who had participated in the Nutrition at the Centre (N@C) project effectively demonstrated better EBF practices than did those who did not participate, we documented the processes of this nutritional intervention in Benin. Methods: This study was a cross-sectional design comparing the intervention group, namely, the Village Saving and Loan Association (VSLA-N@C), to the control group. The N@C project was an educational intervention based on behavioural and social changes related to nutrition. Through VSLA groups installed in communities, mothers were connected to the project; had weekly discussions around the process, benefits and challenges linked to EBF, and advocated during Breastfeeding Week celebrations. The study participants were mothers with children aged 4-5.5 months from the VSLA-N@C group (n = 53) and mothers (n = 50) from non-intervention areas who served as controls. With the deuterium oxide dose-to-mother technique, we quantified human milk intake (HMI) and non-milk oral intake (NMOI) and compared both groups using Student's t-test. A child is considered to be exclusively breastfed if the NMOI is less than 86.6 g/day. Multivariate regression logistics adjusted for VSLA membership, mothers' body mass index, and children's age, weight-forage and weight-for-length, thus enabling us to measure differences in EBF rates. Results: Children of mothers from the VSLA-N@C group consumed significantly more human milk than those of mothers in the control group (900.2 ± 152.5 g/day vs 842.2 ± 188.6 g/day, P = 0.044). Children in the VSLA-N@C group had significantly less non-milk oral intake than did those in the control group (difference: 148.2 g/day, P = 0.000). Therefore, the EBF rate was significantly higher in the VSLA group (38% vs 8%, P < 0.0001), and mothers in VSLAs were 14 times more likely to practise EBF than were those in the control group (adjusted odds ratio [AOR] = 13.9, 95% CI 1.9-116.5, P = 0.015).