Social media have emerged as a promising communication channel for promoting breastfeeding among a new generation of mothers. Yet, there is no published study reporting the effects of a large-scale social media intervention on key breastfeeding-related perceptions, attitudes, and behaviors. As a component of its breastfeeding promotion campaign, the Women, Infants, and Children (WIC) program implemented a 12-month intervention using Facebook and Instagram and subsequently evaluated the outcomes by surveying WIC-participating women (N = 832) twice, immediately before and after the intervention. Based on their level of exposure to the intervention messages, the women were retrospectively classified into two groups, resulting in a two-group (no-low exposure vs. medium-high exposure) quasi-experiment. Women in the medium-high exposure group, in comparison with women in the no-low exposure group, exhibited higher campaign awareness (p < .001), visits to the campaign website (p < .001), and engagement with the website content (p < .001). They also reported more positive breastfeeding attitudes (M = 17.26 vs. M = 16.51, p < .05), self-efficacy (M = 54.48 vs. M = 49.94, p < .01), and social support (M = 27.37 vs. M = 25.11, p < .001). But they did not differ from women in the no-low exposure group in breastfeeding initiation (p > .05) and duration (p > .05). In conclusion, a social media-based intervention resulted in more positive breastfeeding attitudes, higher self-efficacy, and higher perceived social support. Future studies need to investigate the optimal level of intervention message dosage that prompts significant behavioral changes.