Nectarivore sugar preferences and nectar composition in the Cape Floristic Kingdom (southern Africa) differ from trends reported for analogous systems in America and Europe in that sugarbirds and sunbirds show no aversion to sucrose, which is the dominant nectar sugar in many of their food plants. To elucidate the physiological bases (if any) of nectarivore sugar preferences, we determined apparent sugar absorption efficiencies in a passerine endemic to this region, the Cape sugarbird Promerops cafer. Apparent absorption efficiencies for the three major nectar sugars, sucrose, glucose, and fructose, were extremely high (> 99%), as in other specialized avian nectarivores. Xylose, a pentose sugar recently reported in the nectar of some Proteaceae, was absorbed and/or metabolized inefficiently, with a mean of 47.1% of ingested sugar recovered in cloacal fluid. We did not measure the proportions of xylose that were absorbed and/or metabolized. We also compared three methods of estimating absorption efficiency: (1) measurements of total sugar in cloacal fluid with refractometry, without correction for differences between volumes of ingesta and excreta; (2) the same measurements combined with correction for volume differences; and (3) HPLC analyses quantifying individual sugars in cloacal fluid, with correction for volume differences. Refractometry has been frequently used in previous studies. For all sugars except xylose, method 1 yielded results similar to those obtained with method 2, but the convergence was artifactual, and we do not recommend use of this method. Apparent absorption efficiencies calculated with method 2 underestimated true absorption efficiency, because refractometry measures nonsugar solutes, but this error is biologically significant only when efficiencies are low.