Evapotranspiration from tall fescue, perennial ryegrass and zoysiagrass turfs during the summers of 1992-3 was compared to evapotranspiration estimates from an evaporation pan, a black Bellani plate, and several empirical combination models, Actual measurement of turf water use was made with small weighing lysimeters. Soil was maintained at field capacity. Data were collected on 51 dates between June and September. Tall fescue was clipped weekly at 7.6 cm whereas ryegrass and zoysiagrass were clipped 3 times weekly at 2.5 cm, Although differences between the grass species existed, in general the rankings of estimate precision were Bellani plate > evaporation pan > empirical models when compared with measured evapotranspiration rates.