Seawater-atmosphere O(2) exchange rates were experimentally measured in open-top laboratory microcosms. The objective was to establish the relationships between turbulence and oxygen transfer velocity, and thus correct continuously measured day-night changes in dissolved oxygen as estimates of planktonic primary production and respiration. After saturating 15-l sterile seawater microcosms with an oxygen-poor gas mix (4.9% O(2), 95.1% N(2)), the microcosms were left to equilibrate with the atmosphere under different turbulence conditions. The rate of increase in dissolved O(2) was measured at 15-min intervals with polarographic-pulsed electrodes and the corresponding values of the oxygen transfer velocity (the K(O(2)) constant for the different turbulence conditions) were determined. After pooling these and literature data obtained in similar experimental conditions, the relation between epsilon (turbulent kinetic energy dissipation rates) and K(O(2)) was determined. Theoretical K(O(2)) values were also calculated using semi-empirical models in which oxygen transfer velocity (K(O(2))) is related to wind velocity. Theoretical, wind related K(O(2)) values were significantly higher than the experimental ones, and as a consequence overestimate primary production and underestimate respiration rates, even resulting in nocturnal O(2) increase. The magnitude of the differences between experimentally derived and theoretically calculated oxygen transfer velocity, precludes the use of wind-derived equations to calculate K(O(2)) in meso- and microcosms experiments not affected by wind, while the equation obtained relating experimental epsilon and K(O(2)) provides statistically reliable estimations of primary production and respiration.