The primary means by which the oceans serve as a sink for atmospheric carbon dioxide is through the vertical flux of sinking organic carbon. This organic material is derived from primary photosynthetic production in the upper ocean. Its rate of loss to the deep sea, the ''new production'', is rigorously constrained by the upward flux of the limiting nutrient, nitrate. New production estimates have been made by using velocity microstructure measurements to infer nitrate fluxes, where it is assumed that turbulent production (mixing) balances viscous energy dissipation. However, estimates based on a salt finger convection model lead to nitrate fluxes which are as much as an order of magnitude larger than a turbulent mixing model predicts and agree much better with fluxes inferred from biological uptake measurements and tracer studies. Heat flux estimates are also higher, but by less than a factor of 3. Observations of the ''scaled dissipation ratio'' &Ggr;, a quantity that is based on both velocity and temperature microstructure measurements, may provide a means of distinguishing between the two mixing hypotheses. Where both turbulent production and salt fingering contribute to the mixing this distinction is more difficult, but by using measurements of &Ggr; the potential error in the vertical flux estimate due to an improper choice of mixing models is greatly reduced. ¿ American Geophysical Union 1989