Effects of time and storage atmosphere on relaxation properties in persimmon fruit (Diospyros kaki cv Fuyu) were investigated by nuclear magnetic resonance (NMR) imaging during the five weeks following commercial harvest. There were two treatments (n = 4): one in which fruit were hermetically sealed in individual polyethylene bags (modified atmosphere or MA treatment), and another sealed in individual perforated bags (control). Fruit were stored at 7 degrees C for 4 weeks (conditions conducive to development of chilling-injury), before being removed to 20 degrees C for 3 days to simulate poststorage shelf conditions. Every week, and at the end of the shelf-life period, a series of H-1 NMR images of median transverse and longitudinal planes were acquired from each fruit for calculation of spin-lattice (T-1) and spin-spin (T-2) relaxation times. Relaxation times associated with the flesh, vasculature and locules in transverse sections, and flesh from basal, median and distal regions of longitudinal sections, were significantly shorter in MA-treated fruit, although no T-2 treatment difference was noted with vasculature. MA-treated fruit were further distinguished from controls in that all tissues exhibited a sharp increase in T-1 (but not T-2) relaxation between the second and third weeks in cold storage. Within a treatment, T-1 times in flesh and locules were similar (ca. 1120 ms), and shortest in vascular tissue (1036 ms). T-2 was a more sensitive indicator of tissue type (mean values of 71, 101 and 116 ms in the vasculature, flesh and locules of MA-treated fruit, respectively), than T-1. However the relative ranking of flesh and locule tissues was time-dependent. There was no gradient in relaxation times between basal, median and distal regions of the fruit. Incipient stages of chilling-injury development were unable to be identified either by visual inspection of time-course images, or variation in relaxation properties. Our observations do indicate, however, that relaxation properties in fruit were strongly influenced by MA conditions. Increased concentrations of soluble metabolites arising from reduced respiration in MA-treated fruit is a possible mechanism consistent with these results, but not removal of a paramagnetic species like O-2, where T-1 values would be predicted to increase.