The purpose of the experimental development of broad bean fertilisation consultancy was examine the impact of N, P and K supply on broad bean yield, protein content and protein yield, as well as to determine the N, P and K supply limit values on chernozem meadow soil on various nutrient supply levels in a long-term fertilisation experiment. The long-term fertilisation experiment was established in 1989 on deeply calcareous chernozem meadow soil, using 4–4 N, P and K supply levels and a whole treatment combination with 64 treatments. This paper presents the results of experiments performed between 1998–2002. The following consequences can be drawn:

1. Depending on the water supply level of the growing season, the yield of broad bean was between 1.72–3.01 t ha-1 without N fertilisation on chernozem meadow soil with humus content between 2.8–3.2% and satisfactory N supply properties. Yield significantly increased by 0.23–0.45 t ha-1 in each experimental yield until reaching 76–85 kg ha-1 NO3-N in the 0–60 cm layer of the soil resulted from 80 kg ha-1 N fertilisation in the majority of years. No further significant yield increase was experienced on higher N supply levels.

2. During the experiment years, the yield of broad bean increased until 195 mg kg-1 AL-P2O5 supply level in the soil. Broad bean yield did not show any significant change in the soil’s AL-K2O supply level range between 229–490 mg kg-1 in either year.

3. Nutrient interaction was observed between N and P supply levels when the excess N supply and higher P supply resulted in yield decrease or the yield increase effect of P fertilisation was not elicited.

4. Crop year and climatic impacts affect the protein content and protein yield to a larger extent than nutrient supply. In the majority of the experimental years the raw protein content of broad bean increased by nearly 1% as a result of 80 kg ha-1 N fertilisation, when the NO3-N content of the 0–60 cm layer of the soil was between 48–79 kg ha-1. Protein yield significantly increased each year until the 76–85 kg ha-1 NO3-N of the 0–60 cm layer of the soil before sowing, which resulted from 80 kg ha-1 N fertilisation in the majority of years. N supply levels above this value did not result in any significant raw protein content and raw protein yield increase.

5. P supply significantly influences the raw protein content of broad bean in only two out of the four experimental years. These impacts showed an inverse tendency, depending on the given crop year.