The purpose of this study was to test the validity and utility of distinguishing symptoms of anxiety from those of depression and grief in recently spousally bereaved elders. We also examined pathways from baseline (six months or less post-spousal death) to follow-up (12 and 18 months post-death) levels of anxiety, depression and grief-related symptoms. Baseline and follow-up data were available from 56 recently widowed elderly subjects recruited for an investigation of physiological changes in bereavement. Confirmatory factor analyses indicated that a model in which anxiety was specified as a third factor, apart from depression and grief factors, fit the data well and significantly better than either the one or two factor models. Path analyses revealed that both baseline severity of grief and anxiety had significant lagged effects and predicted follow-up severity of depression. Symptoms of anxiety appeared distinct from those of depression and grief, and the anxiety, depression and grief factors differentially predicted subsequent symptomatology. These findings suggest a need for more specific identification and treatment of anxiety, depression and grief symptoms within the context of late-life spousal bereavement.