The total longshore sediment transport rate in the surf zone was measured at a temporary groin installed at Indian Rocks Beach, west central Florida. Approximate mass balance between updrift accumulation and downdrift erosion, which served as an indicator of reliability, was obtained in four of the six runs, which were subjected to further analysis. Magnitudes of three of the four transport rates were considerably lower than predictions by the Coastal Engineering Research Center formula calculated based on contemporaneous measurements of the breaking waves. Error analysis indicated that the determination of the longshore wave-energy flux factor carried a 22–48% maximum uncertainty, and the measured volume-change rate carried a 22–43% uncertainty. The combined uncertainties produce a 40–90% maximum uncertainty in determination of the empirical transport coefficient K appearing in the Coastal Engineering Research Center formula. Comparable or greater uncertainty in K-values probably exists in the total database available for calibrating predictive formulas. The range in values of K in these measurements cannot be explained by measurement error or uncertainty. Therefore, it is concluded that K is not a constant and that other factors may enter, such as breaker type, turbulence intensity, and threshold for sediment transport.