Gradual changes in soil productivity can only be measured by long-term monitoring or field experiments. Long-term agro-ecosystem experiments help evaluation of the cropping system effects on soil quality, sustainability of these systems and provide early warning indication of threats to future crop production (Richter et al. 2007). Fertilisation is one of the most important intensification factors significantly effecting crop yield and quality (Černý et al. 2010). Edmeades (2003) contended that nutrient inputs can increase crop production by 300-400% and are essential in maintaining long-term sustainability. A long-term Broadbalk wheat experiment confirmed a strong yield response to nitrogen fertilisation with the curvilinear character of nitrogen (N) response curve (Brentrup et al. 2004). Jaakkola and YliHalla (2008) reported a significant decrease in cereal yields after 30 years with N fertilisation not balanced with phosphorus (P) and potassium (K) fertilisers. Moreover, yield and its response to fertilisers depend on preceding crops, and the results from a Hungarian long-term experiment (Berzsenyi et al. 2000) showed that yields of wheat cropped in monoculture were always lower than in crop rotation. Simultaneously, significantly higher yields were obtained at high NPK fertilisation rates, especially in rotations with a high proportion of cereals. The combined application of organic manure and inorganic fertiliser was an The influence of mineral fertilisers, liming, farmyard manure and sowing rate on the winter wheat grain yields was studied in a long-term field experiment at 4 sites under different soil and climatic conditions in the Czech Republic. A total of 135 partial fraction-factorial experiments were performed between 1980 and 2013 and evaluated using a statistical model with linear and quadratic terms for each factor. Yield trends demonstrated remarkable influence of fertilisation at two sites of lower starting productivity. Here, grain yields increased by 50% and 25% since the trial commencement, while the rate of yield increase was low at more productive sites. Yields were the most frequently influenced by nitrogen (N) fertilisation, uniformly at all sites. N response curves were strongly curvilinear, but these differed between sites and were affected by preceding crops. The relative frequency of statistically significant influences decreased in the following order: N (significant at α < 0.05 in 89% of all partial trials) > sowing rate (29%) > phosphorus (22%) > farmyard manure (15%) > potassium (12%) > liming (8%). This order and the frequencies of these influences are discussed with regard to relevant site and soil conditions.