Ventilated chambers have been used to measure transpiration rates of various Eucalyptus species. However the microclimate within a chamber differs from natural conditions, leading to possible bias in the measurements. The validity of transpiration rates measured by ventilated chambers was assessed using two techniques which could be applied both with and without the chamber in place. The Penman-Monteith equation, applied to single leaves, provided one set of transpiration estimates for both chamber positions. A second set was obtained using a leaf energy budget method which utilized paired leaves, one of which was prevented from transpiring, while the other behaved normally. Temperature differences between the two leaves and boundary layer conductances measured in situ were used to estimate transpiration rates for normal leaves through solution of leaf energy budget equations. Transpiration rates within the ventilated chamber were reduced by up to 30% around midday relative to undisturbed conditions, according to estimates using the Penman-Monteith equation. Enhanced transpiration during late afternoon was also predicted. Diminished transpiration rates within the ventilated chamber were caused by a marked reduction in available energy when the chamber walls were raised. Recommendations are presented for techniques to measure net radiation absorbed by leaves within a tree canopy. Leaf energy budget measurements on 9–12 paired leaves did not confirm predictions from the Penman-Monteith equation that transpiration rates within the ventilated chamber were lower than under natural conditions. Midday transpiration measurements from leaf energy budgets were ∼ 25% higher than Penman-Monteith predictions when the ventilated chamber was in position and ∼ 25% lower when it was down, thus suggesting little change in transpiration rates with chamber position. Small differences in transpiration rates between trees cannot readily be resolved by any of the estimation techniques used in this study because of inherent experimental uncertainties in each method.