The effects of surface roughness on gas turbine performance are reviewed based on publications in the open literature over the past 60 years. Empirical roughness correlations routinely employed for drag and heat transfer estimates are summarized and found wanting. No single correlation appears to capture all of the relevant physics for both engineered and service-related (e.g., wear or environmentally induced) roughness. Roughness influences engine performance by causing earlier boundary layer transition, increased boundary layer momentum loss (i.e., thickness), and/or flow separation. Roughness effects in the compressor and turbine are dependent on Reynolds number, roughness size, and to a lesser extent Mach number. At low Re, roughness can eliminate laminar separation bubbles (thus reducing loss) while at high Re (when the boundary layer is already turbulent), roughness can thicken the boundary layer to the point of separation (thus increasing loss). In the turbine, roughness has the added effect of augmenting convective heat transfer. While this is desirable in an internal turbine coolant channel, it is clearly undesirable on the external turbine surface. Recent advances in roughness modeling for computational fluid dynamics are also reviewed. The conclusion remains that considerable research is yet necessary to fully understand the role of roughness in gas turbines.