Demineralized bone matrix (DBM) is a widely used bone graft material that derives its osteoinductive potential from matrix-associated bone morphogenetic proteins (BMPs). Prior investigations have shown that the osteoinductive potential can vary widely, with influence from both donor and processing sources. Although it is plausible that donor variance in the BMP profile can be an important consideration, the few published studies available have given inconsistent and incomplete information about this. The goal was to (1) characterize the variance of BMP-2, BMP-4, and BMP-7 in fully demineralized DBM derived from 20 appropriately screened (Food and Drug Administration and the American Association of Tissue Banks criteria) donors (male and female, 17-65 years) and (2) using literature review, infer the potential for this to be an important source of variability in graft function. BMPs were extracted with 4 M guanidine hydrochloride, and levels of BMP-2, BMP-4, and BMP-7 were measured using enzyme-linked immunosorbent assay. Measured levels were as follows: BMP-2 = 21.4 +/- 12.0 ng/g DBM, BMP-4 = 5.45 +/- 2.04 ng/g DBM, and BMP-7 = 84.1 +/- 34.4 ng/g DBM, which were significantly different (P < 0.05). There was a positive linear correlation between BMP-2 and BMP-7 (P = 0.0227). DBM derived from female donors had significantly greater concentrations of BMP-2 and BMP-7 than did that derived from male donors (P = 0.0257 and 0.0245, respectively). There was no significant correlation between donor age and the levels of any of the measured BMPs. The magnitude of variance of BMP profile appears to reasonably well correspond to the variance in osteoinductive potential cited by others, suggesting the possibility of using this as a method of donor screening.