Several cross-sectional studies have shown the ability of the TBS to discriminate between those with and without fractures in European populations. The aim of this study was to assess the ability of TBS to discriminate between those with and without fractures in a large female Caucasian population in the USA. This was a case-control study of 2,165 Caucasian American women aged 40 and older. Patients with illness or taking medications known to affect bone metabolism were excluded. Those in the fracture group (n = 289) had at least one low-energy fracture. BMD was measured at L1-L4, TBS calculated directly from the same DXA image. Descriptive statistics and inferential tests for difference were used. Univariate and multivariate logistic regression models were created to investigate possible association between independent variables and the status of fracture. Odds ratios per standard deviation decrease (OR) and areas under the ROC curve were calculated for discriminating parameters. Weak correlations were observed between TBS and BMD and between TBS and BMI (r = 0.33 and -0.17, respectively, p < 0.01). Mean age, weight, BMD and TBS were significantly different between control and fracture groups (all p ≤ 0.05), whereas no difference was noted for BMI or height. After adjusting for age, weight, BMD, smoking, and maternal and family history of fracture, TBS (but not BMD) remained a significant predictor of fracture: OR 1.28[1.13-1.46] even after adjustment. In a US female population, TBS again was able to discriminate between those with and those without fractures, even after adjusting for other clinical risk factors.