During the third Soviet-American Gas and Aerosol (SAGA 3) expedition to the central Pacific in February to March 1990 we observed C1--C5 nonmethane hydrocarbons (NMHCs) both in the atmosphere and in the ocean. Atmospheric NMHCs project strongly on two factors: those NMHCs with lifetimes over 1 week (ethane, ethyne, propane, cyclopropane) display north-south latitudinal gradients whose magnitudes are proportional to their photochemical loss rates, which is consistent with a northern hemispheric, continental source, while those NMHCs with lifetimes under 1 week (all alkenes and pentane) do not display latitudinal gradients but do vary in common in a more complicated manner, consistent with a heterogeneous, marine source. Data taken from a sea water equilibrator shows that alkene concentrations vary strongly with wind speed, supporting previous conclusions that sea-air evasion is a major loss term for mixed layer alkenes. A reasonable balance exists between sea-air fluxes and atmospheric column removal for some alkenes (ethene, propene, 1-butene, and 1-pentene), while the remainder are out of balance, with partial pressures insufficient to support atmospheric observations. The total NMHC sea-air flux is about 0.8 μmole m-2 d-1. Both our air and water concentrations fall near the low end of previously observed ranges. It is argued that some previous higher atmospheric observations may reflect systematic problems with canister sampling. During SAGA 3 the role of NMHCs in the remote atmospheric OH budge was secondary though important, accounting (in conjunction with NMHC oxidation products) for 10% to 20% of all OH removal during baseline periods and a substantially larger fraction during episodes of increased NMHC mixing ratios. ¿ American Geophysical Union 1993