We examined the hypothesis that thrombophilia, hypofibrinolysis, and the endothelial nitric oxide synthase (eNOS) T-786C polymorphism are common, potentially treatable, and similar pathophysiologic causes of multifocal (three sites or more) and unifocal (single-site) osteonecrosis. We prospectively evaluated twenty-six consecutively referred adults with multifocal osteonecrosis, who included thirteen with idiopathic multifocal osteonecrosis and thirteen with secondary multifocal osteonecrosis (resulting from steroid therapy in ten and alcoholism in three). We compared these patients with race, sex, and age-matched normal control subjects and with patients with idiopathic unifocal and secondary unifocal osteonecrosis, respectively. Using polymerase chain reaction and serologic measures, we studied thrombophilic and hypofibrinolytic mutations and the eNOS T-786C polymorphism. The total number of polymerase chain reaction and serologic thrombophilic-hypofibrinolytic abnormalities and the eNOS T-786C polymorphism did not differ between patients with idiopathic (p > 0.5) or secondary (p > 0.5) multifocal and unifocal osteonecrosis. The frequency of low free protein-S levels (<66%) in patients with secondary multifocal osteonecrosis (four of eleven patients) was higher than that in the control subjects (one of fifty-nine) (risk ratio = 21.5; 95% confidence interval, 2.6 to 174; p = 0.0016, Benjamini-Hochberg adjusted p [Bp] = 0.004). Factor-V Leiden heterozygosity was present in two of thirteen patients with secondary multifocal osteonecrosis compared with none of sixty-four control subjects (p = 0.027, Bp = 0.008). For eleven patients with secondary multifocal osteonecrosis, the eNOS T-786C polymorphism was present in nine of twenty-two alleles compared with eight of forty-four alleles in twenty-two normal control subjects (risk ratio = 2.3; 95% confidence interval, 1.0 to 5.0; p = 0.047, Bp = 0.016). The frequency of homocystinemia (>13.5 mumol/L) was higher in patients with idiopathic multifocal osteonecrosis (two of thirteen patients) than in normal controls (none of fifty-one) (p = 0.039, Bp = 0.004). A high level of factor VIII (>150%) was seen in four of eight patients with idiopathic multifocal osteonecrosis and in seven of forty-eight normal controls (risk ratio = 3.4; 95% confidence interval, 1.3 to 9.1; p = 0.04, Bp = 0.008). The eNOS T-786C mutant allele was present in seven of twelve alleles in the six patients with idiopathic multifocal osteonecrosis who were tested, compared with twenty-five of 108 alleles in fifty-four control subjects (risk ratio = 2.5; 95% confidence interval, 1.4 to 4.5; p = 0.015, Bp = 0.008). Limited by the small numbers of patients with multifocal osteonecrosis, this exploratory study suggested that thrombophilia was associated with both idiopathic multifocal osteonecrosis and secondary multifocal osteonecrosis, as was the eNOS T-786C polymorphism. Multifocal and unifocal osteonecrosis are similarly associated with thrombophilia, hypofibrinolysis, and the eNOS T-786C polymorphism, which are potentially treatable pathophysiologic conditions, requiring further study.