Objective To test the hypothesis that women with epilepsy come from families with poor cohesion and adaptability to the disease, compared with healthy women. Design Cross-sectional study. Setting A retrospective study was done on 263 women from a neurology outpatient department and a family medicine unit from the Mexican Social Security Institute in Sonora (Mexico) between 2010 and 2011. Participants Were 82 women with epilepsy (cases), and 181 healthy women (control group). Age and area of residence was utilized to pair the sample (1:2,2). Measurements Univariate analysis was performed for socio-demographic variables, family type, socio-economic status, level of cohesion and family adaptability based on FACES III. A logistic regression analysis was performed for those variables that were associated with functionality of family in women with and without epilepsy. Results A family dysfunction were found in 22% of epileptic women (OR = 2.91 [2.17, 3.89], p=.0001). Univariate analysis suggested the presence of epilepsy and of family dysfunction associated with age, disease more than 15 years of evolution, and family type (rural, urban and traditional). The logistic regression model confirmed an association only for presence of epilepsy (OR = 7.30 [4.00, 13.33], p=.0001). Conclusion The study answers the hypothesis that women with epilepsy live in families with greater psychosocial impairment, manifested by deficiencies in cohesion and adaptability to the disease, compared with healthy women. © 2013 Elsevier España, S.L. All rights reserved.