Only a single small study (Bullock, L. & McFarlane, American Journal of Nursing, 1989) has clearly defined the increased risk of low birth weight in women abused during pregnancy. Numerous methodologic problems such as limited descriptions of severity, locus and treatment given for injuries, absence of corroboration with specific neonatal findings, and unreliable inferences of cause and effect combined with false variables, limit the extent to which studies can inform clinical practice and guide future research. These problems as a whole stem from the multifactorial nature of family violence.