Uptake and assimilation of nutrients is essential to the productivity of coral reefs living in high-light, low-nutrient waters. Nutrient uptake rates by coral reef communities have been hypothesized to be limited by rates of mass transfer across a concentration boundary layer adjacent to the surfaces of reef benthic autotrophs. Maximum rates of phosphate uptake by a reef flat community were estimated based on the attenuation in wave heights across the community and the rate of frictional dissipation these changes represented. Time-averaged flow speeds varied between 6 and 23 cm/s across the entire study area and over all three days. This was largely due to differences in significant wave heights outside Kaneohe Bay on all three days as well as the attenuation of wave heights across the study area. Differences in wave energy fluxes calculated from first-order cnoidal wave theory were used along with near-bottom orbital velocities to estimate a bottom friction coefficient (cf) of 0.22 +/-0.03 for the reef flat. This value compares well with other estimates of wave friction factors for other reefs (0.28 +/-0.05 and 0.15 +/-0.02). This value of cf, along with other parameters derived from wave height and flow data, was used to estimate phosphate mass transfer coefficients (Sp) across the study area by application of a mass transfer relationship used in prior experimental studies. The dissolution of artificial plaster forms in both a re-circulating flume and on the Kaneohe Bay Barrier Reef flat indicated that rates of nutrient mass transfer are 1.4-2.0 times higher under oscillatory flow than under steady flow. Plaster dissolution was used as an experimental analog for describing maximum nutrient uptake rates given that both mass transfer processes are characterized by constant boundary conditions and high Schmidt numbers (Sc > 500). Taking into account enhancement of mass transfer under wave-dominated, oscillatory flow, SP varied between 5 and 10 m/day across the study area and on all days. Significant wave heights just offshore of Kaneohe Bay on one of the sampling days was close the to the mean significant wave height over a two-year period for which significant wave heights were measured (1.9 m versus 2.0 m). Therefore, spatially averaged data from this day was used to get 8 +/-2 m/day as a representative average value for Sp for the entire reef flat community and over an annual cycle. This value is similar in magnitude to an annually and spatially averaged net phosphate uptake rate coefficient of 9 m/day for the Kaneohe Bay Barrier Reef flat measured by Atkinson (1987). Given typical phosphate concentrations of ~0.1 micro-M across reef flat community studied here, rates of gross and net phosphate uptake are 0.8 +/-0.2 and 0.9 mmol P m^-2 day^-1, respectively. These results indicate that 1) rates of phosphate uptake by this reef flat community operate near the physical limits allowed by mass transfer, and 2) most of the phosphate which is lost from this community is likely exported in particulate form rather than excreted as dissolved phosphate. Scaling maximum phosphate uptake rates by the average C:P of benthic autotrophic tissue (640:1 to 790:1) and comparing these rates with prior estimates of gross primary production indicate that this reef flat community fixes 2-3 times more carbon than can be used for by net primary production. This suggests that net primary production in shallow coral reef communities is regulated by nutrient mass transfer from the water column.