Objective: This study identifies factors associated with child protective services (CPS) involvement among current and former welfare recipients after welfare reform legislation was passed in the US in 1996. Method: Data come from the Women's Employment Study, a longitudinal study of randomly selected welfare recipients living in a Michigan city in 1997 (N = 541). In order to identify risk factors for CPS involvement among current and former welfare recipients, multinomial logit analyses with 29 independent variables were employed on a trichotomous dependent variable: no CPS involvement, investigation only, and supervision by CPS after investigation. Results: The relationship between work and involvement with CPS differs by work experience prior to welfare reform. As the percentage of months working after welfare reform increased, the risk of being investigated by CPS declined among those with prior work experience but the risk increased among those without prior work experience. However, work variables were not significant predictors of supervision by CPS after an initial investigation. Further, race, cohabitation, childhood welfare receipt, having a learning disability, having a large number of children, being newly divorced, living in a high problem neighborhood, and being convicted of a crime were associated with one's probability of being either investigated or supervised by CPS. Conclusions: These findings suggest that employment could have increased the stress levels of current or former welfare recipients without prior work experience to the point where they were prone to minor child rearing mistakes that resulted in a CPS investigation, but were not severe enough to warrant opening the case for supervision. Supports should be provided to welfare mothers who are prone to involvement with CPS; expansions in the childcare subsidy and a reduction or delay in work requirements might also help these families.