Spatial and temporal variations of surface albedo on Haut Glacier d'Arolla, Switzerland, during the 1993 and 1994 ablation seasons are described. Correlation and regression analyses are used to explain the albedo variations in terms of independent meteorological and surface property variables. Parameterizations are developed which allow estimation of albedo variation in surface energy-balance models. Snow albedo is best estimated from accumulated daily maximum temperatures since snowfall. On "deep" snow (≥0.5 cm w.e. depth) a logarithmic function is used, while on "shallow" snow (<0.5 cm w.e. depth) an exponential function is used to enable the albedo to decay to the underlying ice or debris albedo. The transition from "deep" to "shallow" snow is calculated as a function of decreasing snow depth (combined r2 = 0.65). This new parameterization performs better than earlier schemes because accumulated daily maximum temperatures since snowfall correlate strongly with snow grain-size and impurity concentration, the main physical controls on snow albedo. Ice albedo may be parameterized by its relationship to elevation (r2 = 0.28), but this approach results in only a small improvement over the assumption of a constant mean ice albedo.