Ultrasonic measures of bone have been available for clinical research purposes for nearly 10 years, yet there still seems to be a need to compare ultrasound with the accepted gold standard of densitometry. Recently there have been published reports showing that ultrasound measures are associated with both appendicular and hip fracture, in particular after adjustment for densitometry measures. We present here a comparison between speed of sound through the patella and forearm bone densitometry, using their association with prevalent vertebral fractures in a population-based study of women and men. The prospective phase of the Saunders Bone Quality Study includes 1401 women and men who had baseline spine radiographs, patellar ultrasound, and forearm densitometry measurements. Multivariate forward logistic regression was used to determine the age-adjusted odds of vertebral fracture, the number of fractures, and the severity of these fractures, when patellar ultrasound and each of four forearm densitometry measures were entered into the model. Age is the most important factor associated with vertebral fractures, their number, and severity for women, while age is not significantly related to vertebral fractures for men. Of the bone status measures, patellar ultrasound entered the logistic regression models more consistently than any other measure except ulnar bone mineral density for women. The ultrasound measure entered every model for men. We conclude that patellar ultrasound velocity is more consistently associated with the odds of vertebral fractures than radius bone mineral content.