The empirical data used in the present research cover the results of winter wheat cultivar experiments carried out over 1971-1995 and the weather observations reported by local meteorological stations in the north-eastern Poland. The study analysed the relations between winter wheat yielding and meteorological variables, including the yield linear trend, as well as relations between the beginning dates of full development stages of plants and the number of days at a given stage and the temperature and precipitation. Having defined the standard fitting method (R², R²_adj, E), each multiple regression model was verified with Cross Validation (CV) test, the LOO version, to obtain an independent evaluation method. The CV test showed an effective tool as it disqualified almost half of the equations. The significant effects showed a greater impact of weather conditions on plant development than on plant yield.

29-36