This article is Part I of a set of papers addressing mixing in a highly stratified estuary. Measurements of interfacial turbulence were made in Columbia River estuary under conditions of moderate river flow and neap tides. A flux correlation method was used to determine buoyancy fluxes, and fits to theoretical velocity and temperature variance spectra were used to measure turbulent kinetic and potential energy dissipation rates. At the measurement site in the interior of the estuary, mixing along the top of the salt wedge is found to occur only on ebb, during periods when the internal Froude number is supercritical. The wedge cannot, therefore, be in a quasiequilibrium state, and vertical mixing plays a dominant role in its ebb retreat. The turbulent Froude number Fr_T, which represents the ratio of effects of shear and stratification on the mixing, was found to be close to Fr_T ≈ 1 during mixing episodes. This suggests an interplay between shear and buoyancy effects that corresponds to a condition of maximum mixing efficiency Γ. Γ is measured using both direct calculations of the buoyancy and dissipation terms in the turbulent kinetic energy (TKE) balance (this estimate of Γ is referred to as Γ₀), and by estimating the buoyancy term B by its relationship to χ_T, the dissipation of temperature variance (this estimate of Γ is referred to as Γ_d). Averages over several mixing events give Γ₀ ≈ 0.22 and Γ_d ≈ 0.36. Sources of bias in these measurements result from (1) approximations to the full TKE and available potential energy (APE) balances, (2) empirical constants used in fits of measured spectra in inertial subranges, (3) frame motion, and (4) high-pass filtering. The sign of the biases suggest that 0.22 < Γ < 0.36. Flood-ebb asymmetry in mixing also results in a time dependent force balance (discussed further by Kay and Jay [2003] ). This asymmetry is analyzed using a new method to determine profiles of the momentum balance during ebb mixing events.