A study of the water balance of the upper River Ray basin reveals that calculated soil moisture storage exceeds measured storage in winter and falls well below it in summer. The systematic seasonal error is of the order ±50 mm. Several reasons have been proposed which may account for some local water balance anomalies. However, the reported discrepancies are characteristic of a much broader area, and are attributed by Edwards and Rodda (1970) partly to the omission of a soil heat storage term (G) from the Penman equation. The present analysis tests this hypothesis.