Formula milk is a good medium for bacterial growth, so contamination is easy, especially if the preparation and administration pay less attention to antiseptic aspects. This type of research is quantitative using analytic survey methods with cross sectional approach. sampling using accidental sampling technique that is accidental sampling is done by taking cases or respondents who happen to be there or available in a place in accordance with the context of the study as many as 64 respondents. The results of the analysis of respondents with higher education there were 25.49% of babies whose diarrhea, while low education there were 69.23% whose babies had diarrhea, p value = 0.007 was smaller than α = 0.05, indicating that there was a significant relationship between mother's education and diarrhea occurrences. there were 7.14% of respondents who knew how to give formula milk whose babies had diarrhea. while respondents who did not know how to give formula milk there were 83.63% whose babies had diarrhea. The results of the analysis showed that p value = 0,000 is smaller than α = 0.05, indicating that there is a significant relationship between maternal knowledge and the incidence of diarrhea. Respondents who did not use formula milk with diarrhea were 0%. Those who use formula milk who have diarrhea are 42.30%. The analysis results obtained p value = 0.005 smaller than α = 0.05 indicating that there is a significant relationship between formula feeding and the incidence of diarrhea.