A photoperiodic effective daily temperature is proposed as an alternative version of the Thornthwaite scheme for estimating daily evapotranspiration rates. Extensive lysimetric data sets from two contrasting climatic regions were used to test the proposed version. The well-known gross underprediction of the original Thornthwaite approach for arid climates became a slight overprediction under the new version, but it was equivalent to that obtained with the well-recommended Penman–Monteith FAO-56 (PM-56) parameterization scheme. The PM-56 schemes performed well for both environments.