More magnesium, on average, is removed annually from agricultural soils in Finland in yields (10 kg/ha) and through leaching (20kg/ha) than is replaced as fertilizers (4 kg/ha), manure (7 kg/ha) and as wet or dry depositions from the atmosphere (1 kg/ha). The amount of magnesium which is applied in assosiation with liming agents (at the most 25 kg/ha) has a decisie effect on the magnesium balance of the soils, although liming itself reduces those magnesium reserves of the soil which are extractable in neutral ammonium acetate(1 M) or in calcium chloride (0,01 M). The effect of two amounts of nitrogen and potassium fertilizers on the spring cereal and cultivated ley yields, on the magnesium uptake and nutrient contents of the yields, and the effect of liming on the magnesium status of the soil was studied using field, pot and incubation experiments in order to gain an estimate of requirement for magnesium fertilization. The results obtained with magnesium sulphate fertilizer in the same experiments were also used as an indicator of requirement for magnesium fertilization. The ammonium acetate (1 M, pH 7) extractable magnesium in the soil appeared to be the most important source of magnesium for the plants and the best indicator of requirement for magnesium fertilization. For the intensive cultivation of grassland crops, the soil should contain about 15 mg/100g of this type of magnesium. In pot experiments, the plants took up only small amounts of not extractable magnesium. Increasing the nitrogen fertilizer dosage (pot experiments, 4,5 l soil: N1=1500 mg, N2=3000mg N per year, field experiments: N1=50 kg/ha, N2=100 kg/ha N per year) generally brought about an increase in the magnesium uptake and in the magnesium content of the grasses. When the experiments were carried out using coarse mineral soils, the magnesium uptake and magnesium content of the plants decreased during the second and third year as the amount of nitrogen fertilizer increased. This was caused by the low magnesium content of the soil (pot experiments: below 12 mg/100 g soil in neutral ammonium acetate extractable magnesium, field experiments: below 100 mg/l soil in acid ammonium acetate extractable magnesium). In these soils, magnesium fertilization (200 mg Mg for 4,5 l soil per year or 57 kg/ha per year) appeared to have a positive effect on the supply of magnesium to the plants. The heavy clay and sandy clays used in the pot experiment did not require magnesium fertilization. The magnesium in silty clays, which contain a high proportion of the silt fraction (0,02—0,2 mm), may be liberated too slowly for intensively cultivated grasses and it may be necessary to give additional magnesium as fertilizer. Increasing the potassium fertilizer level from 60 kg/ha to 240 kg/ha K per year significantly decreased the magnesium content of the plants. Judging by the quality of the crops, a high potassium fertilizer level appeared to increase the requirement for magnesium fertilization. In the case of muddy very finesand the yield and magnesium uptake of timothy decreased with an increase in the amount of potassium applied. Liming (90, 180 or 360 mg/100g soil Ca as CaCO3) decreased the amount of neutral ammonium acetate extractable magnesium, in seven mineral soils out of nine, by 2—24 % in comparison to the magnesium content of unlimed soils. Part of the fertilizer magnesium became not extractable, too.