Title. Work-home interference among nurses: reciprocal relationships with job demands and health. Aims. This paper is a report of a study with three aims: (i) to investigate whether emotional, quantitative and physical demands have a causal, negative impact on nurses' health; (ii) to examine whether work-home interference can explain this effect, by playing a mediating role; and (iii) to test the so-called loss spiral hypothesis claiming that nurses' health problems lead to even higher job demands and more work-home interference over time. Background. While many scholars have thought in terms of the stressor¿work-home interference¿strain model, the validity of a model that includes opposite pathways needs to be tested. Method. A questionnaire was completed twice, with a 1-year time interval by 753 (63·4%) Registered Nurses working in hospitals, 183 (15·4%) working in nursing homes, and 251 (21·1%) working in home care institutions. The first measurement took place between October 2002 and June 2003. Findings. Our findings strongly support the idea of cross-lagged, reciprocal relationships between job demands and general health over time. The reciprocal model with work-home interference as an intervening variable (including reciprocal relationships between job demands, work-home interference and general health) showed a good fit to the data, and proved to be superior to both the causality and reversed causation models. Conclusion. The higher nurses' job demands, the higher is their level of work-home interference and the more likely is a general health deterioration over time, in turn giving rise to higher job demands and work-home interference, which may even aggravate the nurses' general health, and so on.