The effect of sowing date, N fertiliser rate, plant density and genotype on the yield stability of maize was analysed using 15-year data from a 5×4×5-factorial sowing date experiment, 35-year data from a two-factorial N fertilisation experiment and 25-year data from a two-factorial plant density experiment. Stability analysis on the experimental treatments was carried out using the variance and regression methods. Among the variance parameters, the ecovalence (W), the stability variance (σ²) and the yield stability (YS) were calculated. Based on the data of the sowing date experiment the optimum sowing date (Apr. 24) or sowing ten days later (May 5) were found to be the most stable due to the low, non-significant values of the variance parameters and the values close to unity for the regression coefficients (b). Although early sowing (Apr. 14) led to a significantly higher yield than late sowing, the yield stability was poorer for early sowing. In the long-term N fertilisation experiment the variance parameters indicated the least yield fluctuation at N rates of 80 and 160 kg ha-1, though the yield stability (YS) parameter for the 240 kg ha-1 N rate was also above-average. Regression analysis showed that the yield level and yield stability were the same in all environments for the 160 and 240 kg ha-1 N rates. The stability of the 80 kg ha-1 N rate was similar, but the yield level was approx. 1.3 t ha-1 lower. The yield stability of the plant density response of the maize hybrids was different in each maturity group (FAO number). The stable plant density range was broadest (50-90 thousand plants ha-1) in the FAO 200-299 group. As the vegetation period lengthened the stable plant density range narrowed and shifted towards lower plant densities (for the FAO 400-499 and FAO 500-599 maturity groups: 50-70 thousand plants ha-1). The variance and regression parameters of stability analysis both contributed to the characterisation of the stability of the genotypes and cropping systems investigated. It can be concluded from the results that high yields and yield stability are not necessarily mutually exclusive.