In this article we attempt to identify the impact of social effects on the decision to practice excision on girls, based on the methodology used by Bertrand, Luttmer and Mallainathan (2000). We are particularly interested in social determinants, and make use of the heterogeneity of behaviors according to area of residence, ethnicity and religion. We focus on the interaction between the density and the quality of contacts to infer a social network. We use the percentage of individuals of the same ethnic group and religion, living in the same survey area, to measure the quantity of contacts, and the percentage of excised women of the same ethnic group and religion to measure the quality of contacts. To implement our trials, we use data from the Burkina Faso?s Demographic and Health Surveys 2003, which supplies information on the prevalence of female genital mutilation (FGM) and on the characteristics of Burkina Fasan households. Our results show that social pressure is strongly correlated to the decision to practice excision in Burkina Faso households.