Social workers and other mental health professionals have developed and applied a number of typologies of family types in clinical practice, as well as for program planning. Most such taxonomies have emerged from theory and practice, but few are based on more than two or three dimensions or have been rigorously tested empirically. In this study of a representative sample of 625 households from which clients of a child and family mental health and social services agency come, inductive cluster analysis extracted eight discrete household clusters. The household types included (a) high-aversive households, (b) isolated (mostly single-adult) households, (c) multideficit households, (d) families experiencing circumscribed problems, (e) households with family problems but substantial external supports, (f) extremely isolated families, (g) seriously stressed families with significant strengths, and (h) significant family-functioning problem Demographic factors and other stressors associated with these clusters were also examined Clinical and programmatic implications of these findings are presented, along with suggestions for further systematic replications in other settings.