Buoyancy is an important factor in the dynamic behavior of jet diffusion flames. In order to determine the exact role that buoyancy plays, a simple procedure is described for varying in isolation the relative buoyancy force in stationary laboratory jet diffusion flame experiments. This procedure, which is derived from a theoretical model of these flames, merely requires that background pressure be varied while maintaining constant mass flows of fuel and oxidizer into the burner. It is shown that the sole result of these pressure variations in the theoretical model is that the effective gravitational acceleration acting upon the flame varies as the square of the pressure. Comparisons are made between the structure of a low speed laboratory methane/air flame at various pressures and the results of a direct numerical simulation of the same flame with various gravitational accelerations. Similar evolutions in flame structure are observed in both cases.