Two experimental studies of mildly curved (δ ω /R 5%) two-stream mixing layers, using different fluids and covering a wide range of Reynolds numbers, are analyzed. One study was performed in a low-speed water channel at Stanford University, utilizing flow visualization and laser Doppler velocimetry. In this case, the Reynolds number was low (Re δ ω 7400), the initial boundary layers were laminar, and the velocity ratio was 0.5. The other investigation was performed in the NASA Ames Mixing Layer Wind Tunnel, in which the three-dimensional structure and streamwise evolution of curved mixing layers at high Reynolds numbers (Re δ ω 5.7 10 4 ) were studied, using hot-wire anemometry. Mixing layers with velocity ratios of 0.5 and 0.6, and both laminar and turbulent initial boundary layers, were subjected to stabilizing and destabilizing longitudinal curvature (in the Taylor-Gortler sense). In stable and unstable mixing layers originating from laminar boundary layers, well-organized spatially stationary streamwise vorticity was generated, which produced significant spanwise variations in the mean velocity and Reynolds stress distributions. With the initial boundary layers on the splitter plate turbulent, spatially stationary, streamwise vorticity was generated, which produced significant spanwise variations in the mean velocity and Reynolds stress distributions. With the initial boundary layers on the splitter plate turbulent, spatially stationary, streamwise vorticity was not generated in either the stable or the unstable mixing layer. Linear growth was achieved for both initial conditions, but the rate of growth for the unstable case was higher than that of the stable case. Correspondingly, the far-field, spanwise-averaged peak Reynolds stresses were significantly higher for the destabilized cases compared with the stabilized cases, which exhibited levels comparable to, or slightly lower than, those for the straight case. Universal scaling of the streamwise evolution of the vorticity thickness and Reynolds stresses, using the velocity difference parameter, λ, and the momentum thickness of the high-speed boundary layer, θ 1 , was reasonably successful in grouping the straight, stable, and unstable mixing-layer results over the diverse range of conditions.