Fractures associated with severe trauma are generally excluded from estimates of the prevalence of osteoporotic fractures in the community. Because the degree of trauma is difficult to quantitate, low bone mass may contribute to fractures following severe trauma. We ascertained all fractures in a defined population and compared the bone mineral density (BMD) of women who sustained fractures in either "low" or "high" trauma events with the BMD of a random sample of women from the same population. BMD was measured by dual-energy X-ray absorptiometry and expressed as a standardized deviation (Z score) adjusted for age. The BMD Z scores (mean +/- SEM) were reduced in both the low and high trauma groups, respectively: spine-posterior-anterior (-0.50 +/- 0.05 and -0.21 +/- 0.08), spine-lateral (-0.28 +/- 0.06 and -0.19 +/- 0.10), femoral neck (-0.42 +/- 0.04 and -0.26 +/- 0.09), Ward's triangle (-0.44 +/- 0.04 and -0.28 +/- 0.08), trochanter (-0.44 +/- 0.05 and -0.32 +/- 0.08), total body (-0.46 +/- 0.06 and -0.32 +/- 0.08), ultradistal radius (-0.47 +/- 0.05 and -0.42 +/- 0.07), and midradius (-0.52 +/- 0.06 and -0.33 +/- 0.09). Except at the PA spine, the deficits were no smaller in the high trauma group. Compared with the population, the age-adjusted odds ratio for osteoporosis (t-score < -2.5) at one or more scanning sites was 3.1 (95% confidence interval 1.9, 5.0) in the high trauma group and 2.7 (1.9, 3.8) in the low trauma group. The data suggest that the exclusion of high trauma fractures in women over 50 years of age may result in underestimation of the contribution of osteoporosis to fractures in the community. Bone density measurement of women over 50 years of age who sustain fractures may be warranted irrespective of the classification of trauma.