In the Mediterranean environment characterized by hot, dry summers, a hydrologically oriented field experiment on vegetable crops was carried out between 1988 and 1993 at a site near Naples, Italy. The objective of the experiment was to study the impact of saline water on crop yield and soil properties. The research was carried out on a clay loam soil classified as Haplustolls. Irrigation water was applied at concentrations: 0, 1.25, 2.5, 5, 10 g l −1 of NaCl and at three irrigation intervals of 2, 5 and 10 days. The increasing concentrations were obtained by adding NaCl to fresh water. The irrigation treatments (i.e. solute concentration and irrigation intervals) were repeated consistently over the same plots throughout the experiment. Irrigation schedules alternating sodic and fresh water were evaluated using a numerical simulation model (SWAP) and taking into account the impact of sodic water on soil physical properties. Measurements done during the field trials were used to calibrate and validate the numerical simulation model and to illustrate the consequences on salt and water balance of the irrigation schedules considered in the study. The electrical conductivity of the saturated extracted of soil paste (EC e), the structure stability index and the infiltration rate indicated observable changes in soil physical properties between the irrigation treatments 0 and 1% over the duration of the experiment. Over the years there is clear evidence that no significant change in EC e was observed in the plots irrigated with fresh water (0% T 2 and T 10). On the other hand, EC e increased linearly with time for the treatments irrigated with saline water (1% T 2 and T 10). Degradation of soil structure was evident in the observed infiltration rate: <1 mm h −1 in the treatment 1% versus >10 mm h −1 in the treatment 0%. Changes in soil hydrological properties were evaluated by determining the h( θ) and K( h) relationships of undisturbed soil cores for four irrigation treatments and using the van Genuchten parametric model of these relationships. For the 2- and 10-day irrigation frequencies, the h( θ) curve of the 1% treatment had lower values of θ than the 0% treatment at the same pressure head in the range 2.0–3.0 pF. Significant differences were observed between the mean values (for each treatment) of most of van Genuchten’s parameters, particularly between the (0% 2 day) and (1% 2 day) treatments. Prior to the evaluation of irrigation schedules the SWAP model was calibrated and validated against the data available in the 1993 (eggplant crop). Irrigation schedules were compared on the basis of a performance indicator ( S c), which measures the relative change in the amount of adsorbed and dissolved salt over an entire irrigation season.