To characterize tree responses to water deficits in shallow and deep rooted conditions, parameters developed using daily oscillations from continuously measured soil water content and trunk diameter were compared with traditional discrete monitoring of soil and plant water status in lysimeter and field-grown peach trees [Prunus persica (L.) Batsch `O'Henry']. Evaluation occurred during the imposition of deficit irrigation for 21 days followed by full irrigation for 17 days. The maximum daily available soil water content fluctuations (MXAWCF) taken at any of the four monitored root zone depths responded most rapidly to the deficit irrigation. The depth of the MXAWCF increased with time during the deficit irrigation. Differences relative to a fully irrigated control were greater in the lysimeter than the field-grown trees. Minimum daily trunk diameter (MNTD) and maximum daily trunk shrinkage (MDS) responded sooner than midday stem water potential (stem Ψ), predawn or midday leaf water potential (predawn leaf Ψ and leaf Ψ), or photosynthesis (A). Parameters based on trunk diameter monitoring, including maximum daily trunk diameter (MXTD), correlated well with established physiological parameters of tree water status. Statistical analysis of the differences in the measured parameters relative to fully irrigated trees during the first 10 days of deficit irrigation ranked the sensitivity of the parameters in the lysimeter as MXAWCF > MNTD > MDS > MXTD > stem Ψ = A = predawn leaf Ψ = leaf Ψ. Equivalent analysis with the field-grown trees ranked the sensitivity of the parameters as MXAWCF > MNTD > MDS > stem Ψ = leaf Ψ = MXTD = predawn leaf Ψ > A. Following a return to full irrigation in the lysimeter, MDS and all the discrete measurements except A quickly returned to predeficit irrigation levels. Tree recovery in the field-grown trees was slower and incomplete due to inadequate filling of the root zone. Fruit size was significantly reduced in the lysimeter while being minimally affected in the field-grown trees. Parameters only available from continuous monitoring hold promise for improving the precision of irrigation decision-making over the use of discrete measurements.