Phosphate uptake (P-uptake) into coral reef communities has been hypothesized to be mass-transfer limited. One method of demonstrating mass-transfer limitation of P-uptake is to show dependence of P-uptake on water velocity. Water velocity across reef flats varies with tides and swell; thus, we measured P-uptake over the entire reef flat on eight different days, representing a range in water velocities. P-uptake was calculated from changes in P concentration of the water column. Changes in P concentration were measured by water sampling at six sites along a 300-m cross-reef transect while simultaneously measuring water velocity. To smooth the variability in phosphate concentrations, peristaltic pumps were used to get time-integrated water samples for 4–6 h at each site. Water velocities were measured in the middle of the transect using an acoustic Doppler current profiler and were averaged to match the time-integrated water sampling. Depth-averaged cross-reef water velocities were 0.031 ± 0.013 m s −1 (mean ± SD), while the root-mean-square water velocities, accounting for oscillatory flow, averaged 3.3 times higher, 0.101 ± 0.021 m s −1 (mean ± SD). Phosphate decreased along all transects. The first-order rate constant for P-uptake (S) was 8.5 ± 2.4 m d −1 (mean ± SD) and increased linearly with root-mean-square water velocity. The Stanton number derived from oscillatory flow, the ratio of the first-order rate constant for P-uptake to the root-mean-square water velocity ( S / U rms ), was (9.4 ± 1.2) × 10 −4 (mean ± SD). P-uptake ranged from 0.2 to 1.1 mmol P m −2  d −1 , demonstrating that P-uptake is variable on short time scales and is directly related to P concentration and water velocity.