Relationships between midday (Ψmd) and predawn (Ψpd) leaf water potential, stomatal conductance (gs), and net CO2 assimilation rate (A) were determined at different fruit growth stages and for 2 years with different fruit loads in a `Sudanell' peach [Prunus persica (L) Batsch] plot subjected to two regulated deficit irrigation (RDI) strategies plus a control irrigation treatment. A postharvest RDI (PRDI) treatment was irrigated at 0.35 of the control after harvest. The second treatment (SPRDI) applied RDI during Stage II, the lag phase of the fruit growth curve, at 0.5 of the control and postharvest at 0.35 of the control. The control treatment and the PRDI and SPRDI when not receiving RDI were irrigated at 100% of a modified Penman crop water use calculation (ETo) in 1994, a full crop year, and 80% in 1995, a year of nearly zero crop. In 1995, with 80% of the 1994 irrigation rate and no crop, the Ψmd was higher, probably because of the lower crop load, while Ψpd was lower, probably because less water was applied to the soil. The relationship of gs and A with Ψmd during Stage II was steeper than during postharvest. Low Ψmd was not indicative of a depression in gs and A in Stage III. Osmotic leaf water potential at turgor loss (Ψπ0) as derived from pressure-volume curves was more negative during Stage III and postharvest (about -2.9 MPa) than in Stage II (about -2.7 MPa). The Ψmd measurements together with Ψπ0 determinations seemed to be more useful to characterize peach tree water status than Ψpd under soil water deficits because of their better relationship to midday stomatal closure.