Nitrogen (N) and Phosphorus (P) are the main nutrient constraints for grain production and are normally supplied as fertilizers to cropping sequences. Reported responses to secondary nutrients have been scarce, inaccurate and circumscribed to experiments carried out in farmers' plots, where systematic monitoring is difficult or not possible. In particular, this approach does not allow looking for the residual effects or interaction of the applied nutrients. Long-term experiments mitigate those problems due to repetitive nature of the evaluation in a same site that also decreases the experimental error. The objective of this work is to study the direct, residual and cumulative effect of applications of Potassium (K), Magnesium (Mg), Chlorine (Cl), and Sulfur (S) on grain yields of wheat, soybean and maize cropped in sequences conduced in six experimental stations of the Pampean region of Argentina during five years. The experiments began in 2000, in six representative locations of the main grain producing area of Argentina: Manfredi, Paraná, Oliveros, Pergamino, Anguil and Balcarce, from North to South. A wheat/soybean – maize sequence was grown in four of the northernmost locations, and the remaining two were cropped to maize - wheat-oats. Ten treatments, consisting in an incomplete factorial combinations of Potassium (K), Chlorine (Cl), Sulfur (S) and Magnesium (Mg), were applied annually on each cereal crop as different fertilizer sources; this included a check with no fertilizer addition in the five years, and a control with N and P but no secondary nutrient application. The accumulated yields show the different potential for grain production of the sites. After two cycles, only a strong contrast between check treatment (zero nutrient application) and the fertilized plots could be recorded, while the contrast between the control treatment receiving N, P and the treatments with secondary nutrients is weak , marginal and inconsistent across sites, and across nutrients. The average difference between check and control is 1563 kg ha^-1 year^-1 or 44%. Since fertilizer carrying secondary nutrients included different combination of cations (K and Mg) and anions (S and Cl) according to the available fertilizer sources, a stepwise analysis were carried out on each grain crop, overall years, to identify the contribution of each nutrient in the mean grain yield. An overall trend revealed that K appears to influence grain yields in several site - crop combinations and so seeming deficient. Since typical K levels in the experimental sites are ranging from 1 to 4 cmolc kg^-1, we conceive that the additional effect of yield contribution seems to be linked to input of either Cl and or S, which are clearly deficient for some crop-site situations. Potassium, combined with either S or Cl, increased corn and wheat grain yields in four of the six sites, and when combined only with S affected soybean grain yield in three of the four sites. The overall response was 9,4 kg ha^-1 of corn and 4,7 kg ha^-1 of wheat per kg of applied K. The grain response to these nutrients application appeared to enlarge with time. It is concluded that the application of potash in the sulfur or chloride form improves yields of main grains in the evaluated locations in the Argentinean Pampean Region.