The psychometric properties of an adjusted version of the Texas Grief Inventory (Zisook, Devaul & Click, 1982) were evaluated in a sample of 207 women following an early pregnancy loss. Factor structure of the Inventory was examined, and the validity was assessed by comparing the level of grief in women who became pregnant following a miscarriage and those who did not. Factor analysis resulted in a three-factor solution that was theoretically interpretable. The identified subscales were: pure grief, grief-related emotions and perceived adjustment and functioning since miscarriage. Significantly lower levels of grief-related behaviours and feelings were reported in women who became pregnant, which was suggestive of the construct validity of the Inventory. The high reliability coefficient (Cronbach's alpha = .90) obtained with this 17-item version of the Grief Inventory suggests it to be a reliable instrument for the assessment of grief after early miscarriage.