Various studies have analyzed the variables that influence the outcome of family foster care. The research aimed to determine whether distinct profiles of foster placements can be identified on the basis of key variables, seeking at the same time to define a predictive function for the success or failure of foster care. Participants were the foster families of 104 non-relative foster children. A k-means cluster analysis identified three clusters, corresponding to high-, moderate-, and low-risk placements. The variables that formed part of these clusters were behavior problems and impulsivity/inattention in foster children, level of burden in the foster parents, an authoritarian parenting style, and criticism/rejection by the foster parents. A discriminant analysis confirmed the differences between the three clusters and enabled the researchers to create a function for classifying cases in each group. These results can be used to identify at-risk placements and may help to avoid situations that could undermine the foster child’s development. The findings could also be useful in terms of assessing the suitability of foster families, as well for identifying their training needs.