The purpose of the present cross-sectional study was to examine the effects of ambiguity of loss and type of caregiver-to-patient relationship on anticipatory grief, negative physical and psychological outcomes associated with grief, and boundary ambiguity in family caregivers of chronically ill patients. Questionnaires were completed by 23 parents of ill children and 30 spouses of ill mates. Using an original and a revised concept for level of ambiguity, partial support was found for the prediction that parents and spouses in high ambiguity of loss circumstances would report more anticipatory grief than those in low ambiguity ones. Contrary to prediction, a slight but nonsignificant trend occurred for parents and spouses in low ambiguity situations to report more negative physical and psychological effects associated with grief as well. Level of ambiguity was not found to impact boundary ambiguity as had been hypothesized. Spouses reported more boundary ambiguity than parents, regardless of level of ambiguity of the loss. Contrary to prediction that parents would report less anticipatory grief and more negative physical and psychological outcomes than spouses, generally, no significant differences were found between the two groups. However, using the original concept of ambiguity, parents did tend to recall more past grief than spouses. The study highlighted several methodological concerns which impact research on loss and grief, particularly the difficulty involved in recruiting participants with subsequent occurrence of sampling bias, rudimentary status of available measurement tools, and a host of potentially confounding personal and sociodemographic variables. The present study supports a view of the loss which occurs in families dealing with chronic illness as a complex process whose impact on grief, distress, and family upheaval is influenced by multiple factors. Such factors include both the ambiguity of the loss and the type of family relationship involved. Complex research of a longitudinal nature using psychosocial models of illness is needed to better delineate the impact of factors such as these.