Energy dissipation in a western boundary current begins with the conversion of mean potential energy into kinetic energy of “primary” eddies. The rate of this energy conversion is taken to equal the total energy dissipation rate, in analogy with the energy cascade of laboratory turbulence. The growth of primary eddies is due to baroclinic instability, and implies the rising of relatively light fluid, sinking of heavy fluid. An inviscid, nondiffusive model adequately describes this process and yields a quantitative relationship between the dissipation rate and upward-downward motions. In statistically steady flow, an eddy mass transport arises from the correlation of cross-stream velocity and isentropic layer thickness which is proportional to the energy dissipation rate. In this manner, the rate of upwelling of upper thermocline fluid and/or the downwelling of heavier fluid comes to determine the energy dissipation rate. Conversely, a given energy dissipation rate implies upwelling or downwelling of specific intensity. A simple parameterization scheme for eddy mass transport similar to that suggested by Green for the atmospheric jet stream gives results in accord with the meager experimental evidence on western boundary upwelling. In virtue of the connection between upwelling and dissipation, this scheme yields a formula for energy dissipation per unit length of the boundary current. An interesting point is that Rayleigh friction between the fast core of the current and the underlying fluid gives the same cross-stream average dissipation if only the friction coefficient is appropriately chosen. The predicted dissipation rate varies with the bulk properties of the boundary current (e.g., the rise of the thermocline across the current) in a realistic way, and the total dissipation integrated along the boundary current agrees in order of magnitude with the energy input into a subtropical gyre by wind.