From September 1980 through May 1983 a series of nineteen sections of velocity profiles were obtained across the Gulf Stream 200 km northeast of Cape Hatteras. By decomposing the velocity and temperature observations into mean and fluctuating fields in two coordinate systems, geographic (or Eulerian) and 'stream' coordinates, it is shown that at least ⅔ of the eddy kinetic and potential energy is caused by the meandering of a well defined baroclinic front with a structure that is nearly independent of space and time. It is also shown that more than 95% of the kinetic energy of the front can be accounted for by a barotropic and a baroclinic mode with near equipartition between the two.The cross-stream baroclinic, barotropic, and pressure-work terms in the eddy energy production equation are estimated to determine what processes contribute to the rapid growth of meandering after the current leaves the coast. In order of importance, the cross-stream average of the baroclinic conversion term is a factor three larger than the other two. The cross-stream averaged production of eddy energy is, however, clearly too large to be consistent with the observed rate of growth of the meander envelope since it would lead to a doubling of eddy kinetic and potential energy in only 2.1 days or 50 km following the mean flow. It is shown that in the case of the baroclinic conversion term the large cross-stream covariances 〈uT′〉 have a simple geometric interpretation in terms of meander growth (and decay). They represent a down (or up) gradient heat flux that is not actually participating in the conversion processes suggesting that the baroclinic production terms are nearly horizontally nondivergent. Similarly, the pressure-work terms must be very nearly horizontally nondivergent (geostrophy). Thus, estimates of energy conversion rates are bound to be greatly exaggerated unless both horizontal components are included. Furthermore, conclusions about the relative importance of the cross-stream conversion terms to the production of eddy energy depend upon their horizontal divergence being in the same proportions, a very unsatisfactory assumption.A simple kinematic model is used to show that the amount of energy needed to support meander growth is quite small. It is clear that to determine these rates experimentally puts great stress on conventional measurement procedures and suggests that alternative approaches such as paying more attention to boundary or flux conditions might be more rewarding in future studies.