Study objective: To characterize injuries to battered women by compari
ng their location, type, and severity with those of injuries to women
resulting from other mechanisms. Methods: Cross-sectional study of 9,0
57 women between the ages of 19 and 65 years who presented for any rea
son to the emergency departments of 10 hospitals serving inner-city, u
rban, and suburban populations. Results: A total of 280 injured, batte
red women were identified during the study period. About 3.1% (95% con
fidence interval [CI], 2.7% to 3.5%) of all women seen in the ED, and
11.2% (95% CI, 10.0% to 12.4%) of injured women with known mechanisms
of injury, were determined to be positive for battering. Battered wome
n were more likely to be injured in the head, face, neck, thorax, and
abdomen (P<.001) than were women injured by other mechanisms. Twelve s
pecific injury types were identified that occurred more frequently in
battered women. Conclusion: Although battered women experience certain
injury types more frequently than women injured by other mechanisms,
the low positive predictive value of these injuries supports the use o
f universal screening for domestic violence in all injured women.