Bile acid synthesis can be measured as release of14CO2 from [26-14C]cholesterol divided by cholesterol specific activity, but this method has not been validated in human subjects. We made twelve comparisons of this CO2 method to standard isotope dilution in six normal subjects and found a mean discrepancy of 6%. Linear regression analysis of one value with respect to the other revealed a correlation coefficient of 0.83 (P<0.001), a Y-intercept close to zero (-4.98) and a slope close to 1 (1.06), suggesting good correspondence between the two methods. To assess the potential for error arising from use of serum cholesterol to estimate specific activity of cholesterol used for bile acid synthesis, we compared synthesis measured using serum free cholesterol specific activity to that measured using bile cholesterol specific activity, which is known to be near isotopic equilibrium with the precursor pool used for bile acid synthesis. Synthesis calculated in these two ways differed by less than 10%. The data indicate that the CO2 method using either serum or bile cholesterol specific activity provides a valid estimate of bile acid synthesis in man.