The studies included in this review met the following criteria: published in a psychological, medical, sociological journal or book; provided enough statistical information so that the information could be meaningfully analyzed (e.g., conducted statistical analyses comparing abused versus non-abused groups, provided more information than statistical significance levels); employed either a representative community sample or a clinical sample with an appropriate control group; and isolated child psychological abuse specifically. A review of sociodemographic variables found that no sociodemographic risk factors with moderate effect sizes (in the hypothesized directions) emerged. It is likely that very low family income is such a risk factor, but an effect size was not published, nor could it be computed. Regarding child characteristics, Vissing et al. (1991), in their nationally representative sample of 3,346 families, found that child aggression, child delinquency, and child interpersonal problems were significantly associated with the child being psychologically abused. Effect size for these relations could not be calculated. Regarding parent characteristics, several parent risk factors had at least moderate effect sizes for their relation to child psychological abuse: aggression and hostility, neuroticism, relationships with fathers perceived as less caring, and being yelled at daily as a child. In a study by Lesnik-Oberstein et al. (1995), psychologically abusive and non-abusive Dutch mothers were compared on a variety of marital relationship measures. Psychologically abusive mothers reported receiving less affection from their husbands and giving less affection to their husbands than did non-abusive mothers. Marital relationships among abusive mothers also tended to be characterized by greater levels of verbal and physical aggression. Limitations of this review and directions for further research are discussed. 1 table and 20 references