A consequence of climate changes is an increasing frequency of drought, which on average occurs in Croatia every third to fifth year and during the vegetation period it can reduce crop yield significantly. The aim of the research is both to determine crop water requirements in an average and in a dry year and to determine the decline in crop yields in an average and in a dry year. The multi-annual climate data series for a 30-year period, 1981-2010 from station Pazin in central Istria and 1981-2010 from Poreč in western Istria was used. Based on these data, a reference evapotranspiration was calculated for an average and a dry year using the Penman-Monteith method through "Cropwat" software. The crop water requirement for five different crops is determined by soil water balance using the Palmer method (Palmer, 1965), corrected according to Širić and Vidaček (1988), using "Hidrokalk" software. Crop response to the lack of soil water and yield decline were determined according to the method published by Doorenbos and Kassam (1979). Correlation test was used to determine correlation between precipitation and crop yields. In central Istria, water shortage in an average year ranged from 3.4 mm (olives) to 110.7 mm (alfalfa) and yield decline ranged from 1.1% (olives) to 18.6% (alfalfa), while in a dry year water shortage ranged from 43.2 mm (olives) to 229.5 mm (alfalfa) and ranged from 10.3% (olives) to 37.7% (alfalfa). In western Istria, in an average year water shortage ranged from 35.5 mm (olives) to 239.7 mm (alfalfa) and yield decline ranged from 7% (olives) to 40.2% (tomatoes), and in a dry year water shortage ranged from 74.4 mm (olives) to 288.9 mm (alfalfa) and yield decline ranged from 14.4% (olives) to 40.7% (alfalfa). The determined water shortage and reduced yields are sufficient indicators of irrigation requirements in Istria.