CO2 uptake in a nearshore coastal ocean was directly measured using eddy covariance over 27 months in the mid-latitude eastern Pacific. The average annual CO2 uptake during the study was 11.8 mol C m−2yr−1 roughly 25 times greater than the global ocean average. Distinct temporal patterns of CO2 uptake were resolved at diurnal, and seasonal scales. Regression analysis of variance indicates that the major drivers of the observed patterns of uptake are related to photosynthesis, with both photosynthetically active radiation and sea surface temperature explaining more of the variability than wind. Warmer sea surface temperatures were correlated with stronger CO2 uptake, the opposite of what would be expected if the "solubility pump" were in action. Air-sea CO2 flux values calculated using the pCO2 data and a constant k value of 10 cm hr−1 were only 25% of the magnitude of direct eddy covariance measurements. The average k value calculated using eddy covariance CO2 flux and delta pCO2 values was 48.8 cm hr−1. Use of the eddy covariance method to directly measure CO2 exchange at the air-sea interface in coastal waters avoids key uncertainties encountered when exchange is calculated using pCO2 differences and k-wind speed relationships, and it is suggested that the discrepancies that have been observed between eddy covariance CO2 flux measurements and pCO2 models could be the result of inadequate inclusion of complexities in pCO2 models, and that direct measurement via eddy covariance is the preferred method in coastal regions. It is concluded that global kelp and sea grass communities could be taking up as much as 2.1 x 1014 g C (0.21 Pg C) of atmospheric CO2 each year. Therefore, macrophyte ecosystems, which comprise between 0.6-2.2% of the surface area of the ocean could be responsible for as much as 10% of the value currently estimated for total global ocean CO2 uptake