Drawing on the Effort-Recovery (E-R) model, the current study investigated to what extent Work-Home Interference (WHI) mediated the relationship between workload and two indicators of well-being, that is, (a) affective well-being (i.e., work-related negative affect and depressive mood) and (b) subjective health (i.e., health complaints). In Part I of this study, Structural Equation Modeling (SEM) was used to test several competing models of mediation - full, partial, or no mediation - in three homogeneous samples (166 medical residents, 194 child care workers, and 224 bus drivers). In Part II of this study, we cross-validated the best fitting model in an independent heterogeneous sample (1421 Dutch workers). The results provided support for the E-R model in that WHI played a significant role in mediating the impact of workload on workers' well-being. WHI fully mediated the relationship of workload with depressive mood and health complaints, and partially mediated the relationship with work-related negative affect. This differential role of WHI indicates that WHI might play a more crucial mediating role with respect to general (context-free) indicators of well-being than with respect to work-related indicators of well-being. In general, the findings of the current study suggest that workload exerts its negative effects on well-being (at least partly) through a process of spillover of negative load-effects that impede recovery during the non-working hours. © 2002 Elsevier Science (USA). All rights reserved.