Most research into caregiver employment and outcomes conceives of employment as a dichotomous variable—employed or not. This study examines the relationship between work interferences and caregiver burden, well-being, and self-esteem within a modified stress process model. Regression models are employed using a population-based random sample of caregivers. Employment status effects on outcomes for the total sample are estimated, followed by estimations of the effects of work interferences on the same outcomes for the employed subsample. Employment status is unrelated to outcomes in the total sample. In the employed subsample, a 13-item work interferences scale is related to burden but unrelated to well-being and self-esteem. Of the 13 items, only “performance at work was affected because of caring for a care recipient” is associated with all three caregiver outcomes. These findings suggest that subjective assessment of work interferences may play a more important role than does employment status.