Ambivalence has been viewed as an important causal agent in the development
of complicated grief. However, examination of studies commonly cited as su
pporting this belief reveals basic limitations in their methodology and con
clusions. The current study examined associations between several relations
hip predictors (ambivalence, affiliation, and dependence) and both grief-sp
ecific symptoms and depression in two samples of psychiatric outpatients wh
o had experienced loss of significant others. Findings from the first sampl
e (N = 138) were used to test for evidence of cross-validation in the secon
d sample (N = 139). Contrary to traditional belief, ambivalence was inverse
ly related to severity of grief symptoms. In contrast, affiliation and depe
ndence were directly related to severity of grief symptoms. None of the pre
dictors provided evidence of cross-validation in the case of depression. Ex
planations for the fmdings and clinical implications are considered.