The reliabilities of the radiocarbon and uranium series inequilibrium methods of absolute age determination as applied to lacustrine carbonates have been evaluated. Fifty samples of gastropods, chara, tufa, marl, and ostracods from the Pleistocene deposits of Lakes Lahontan and Bonneville were analyzed for C14, U238, U234, Th232, Th230, and Ra226. The results indicate that samples in which Ra226 is at equilibrium with Th230 and where the Th232 concentration is not unusually large, satisfactory agreement between the two methods generally exists. Whereas a common thorium correction is required, evidence is given that for the basins studied the Th230–Th232 ratio has a nearly constant value of 1.70. Where disagreement between the two methods occurs, the lower age appears to be in error. The anomalies appear to reflect U addition (low Th230 age) or contamination with young carbon (low C14 age). Samples beyond the range of radiocarbon yield Th230 ages consistent with stratigraphic information, offering the possibility of extending the absolute chronology of the fluctuation in the size of these lakes back to 200,000 years ago. Independent age estimates based on the disappearance of the excess U234 incorporated into these samples are clearly invalid. Samples with Th230 ages of more than 100,000 years show excesses of U234 comparable to or even greater than those for samples less than 20,000 years old. Support for Attree's (1962) estimate of 75,200 years for the Th280 half-life has been obtained (opposed to the previously accepted estimate of 80,000 years).