Plant age and size, seasonal growth patters and crop load, among other factors, have been reported to decrease the usefulness of trunk diameter variation (TDV) derived indices as water stress indicators in olive trees. Our hypothesis, however, is that indices derived from TDV records in old, big olive trees are sensitive enough to detect levels of water stress in trees of orchards under deficit irrigation that, although severe, are below the threshold for fruit shrivelling. This is of importance for the production of good quality oils, since fruit shrivelling may affect oil quality. The aim of this work was to assess different TDV-derived indices as water stress indicators in 40-year-old �Manzanilla� olive trees with heavy crop load. We derived the maximum daily shrinkage (MDS), daily growth (DG) and daily recovery (DR) from TDV records taken during the 2008 dry season both in well-irrigated FAO trees and in deficit-irrigated RI trees. Measurements of volumetric soil water content (?v), leaf water potential (?l), stomatal conductance (gs), net CO2 assimilation rate (A), water and oil accumulation in the fruits and yield parameters were made for both treatments. The trunks did not grow during the experimental season, either in the FAO or RI trees, likely because of the heavy crop load. Therefore, DG was useless as water stress indicator. For MDS and DR, which were responsive to the increase of the trees� water stress, we calculated the variability, quantified by the coefficient of variation (CV), the signal intensity (SI) and the sensitivity (SI/CV) values. In addition, we derived reference equations for irrigation scheduling from the relationships between MDS values in the FAO trees and main meteorological variables. Values both of SI-MDS and SI-DR were steady until September 9, despite of increasing differences in ?v between treatments from early in the dry season. The ?l vs ?v values showed an outstanding capacity of the RI trees to take up water from the drying soil, and the ?l vs gs values showed a near-isohydric behaviour of those deficit-irrigated trees. These results explain, at least in part, the lack of response of MDS and DR on that period. Both SI-MDS and SI-DR peaked for the first time on September 9, 16 days before the appearance of fruit shrivelling. Our results suggest that using TDV-derived indices as water stress indicators for irrigation scheduling in old olive orchards with medium to low plant densities, i.e. with large root zones, may be useless in case the irrigation strategy is aimed at keeping the soil close to field capacity. Nevertheless, the MDS and DR indices may be useful indicators for the avoidance of fruit shrivelling in deficit irrigated olive orchards for the production of good quality oil. Reliable reference equations for scheduling irrigation with the signal intensity approach were obtained from the regression of MDS values vs the daily maximum values of both the air temperature and the vapour pressure deficit of the air