Abstract Poor sanitation may be associated with growth faltering, but empirical evidence on the association between household sanitation infrastructure and child nutrition has been mixed. This article assessed whether sanitation infrastructure is associated with growth penalties for Roma children. Design This is a Roma nationally representative population-based study. Proxies of child nutritional outcomes included children’s height-for-age Z scores below two standard deviations from the median of WHO’s reference population. Multiple and logistic regressions estimated the association between the type of toilet usage, outcomes, and sociodemographic determinants. Setting Data were from UNICEF’s fifth and sixth Multiple Indicator Cluster Surveys for Serbian Roma settlements. Subjects Children aged 0–24 (n = 1,150) and 25–59 (n = 2,024) months born to women aged 15–48 years were included in this study. Results About 16% of children resided in households with no toilet service, and 23% of children were stunted. Roma children aged 25–59 months in households with no toilet service and in rural areas were more likely to be stunted and of significantly shorter stature than their counterparts. Children up to 24 months of age were not affected by the type of toilet, likely due to protective effects of breastfeeding. Child’s age, number of siblings, and birth spacing negatively influenced growth outcomes, while maternal age, birthweight above 2.5 kg, and being a female mitigated against poor growth outcomes. Conclusions The relationship between the type of toilet and growth outcomes among Roma children varied with age and area: having no toilet and living in rural areas were associated with compromised child growth for children aged 25–59 months. Maternal and child characteristics additionally accounted for children’s growth.