Sixty-seven persons were identified 1 to 2 months after the death or l
ife-threatening illness of their spouse and followed for 25 months. In
take measures included a) a revised Ways of Coping Scale, a structured
assessment of ego defenses, sociodemographic information, and other b
aseline variables. Fifty-six completed follow-up. Outcome measures inc
luded deaths, hospitalization, self-rated health, depressive symptoms,
symptoms of anxiety, and separation distress. In our analyses, bereav
ement was used as a covariate and found to be unrelated to outcome. Lo
w self-ratings on coping by making a change and problem-focused planni
ng predicted higher scores on separation distress at 13 months (p less
than or equal to .05). Participants who used less problem-focused pla
nning were at risk for higher depression scores 13 months after the st
ressful event (p less than or equal to .05). Low ego-defensive work an
d high neurotic ego-defensive ratings predicted high depression scores
at 13 months (p less than or equal to .05). At 25 months, coping by s
elf-blame was inversely related to scores on separation distress (p le
ss than or equal to .05). Coping variables predicted neither scores on
anxiety symptom scales nor the outcomes of hospitalization or death o
ver the 25-month study period. These observations counter some prevail
ing clinical assumptions about coping with a loss and emphasize the va
lue of empirical studies of coping as a mediator of outcome during the
stress of a loss.