The use of x-ray tubes as radiation sources for Compton scatter densitometry (CSD) is evaluated theoretically. Bias correction factors (bcf's) which correct for differential attenuation due to Compton shift and polychromaticity, and differential scattering cross section due to polychromaticity, are defined to facilitate comparison of monochromatic and x-ray tube CSD devices. Computer modeling indicates that a 100-kVp x-ray tube spectrum hardened with 0.64 mm of Pb filtration yields bcf's 0.008 to 0.34% higher than a 70-keV monochromatic source for various samples of nylon, and human muscle and bone. This difference is insignificant because multiple scatter errors are typically 1000% higher than errors corrected by bcf's. The high equivalent activities of x-ray tubes (10 to 80 TBq) and their small focal spots (1.0 to 2.0 mm) allow the long-standing CSD problems of marginal statistical precision, long counting times, and low spatial resolution to be overcome.