According to the. established technique cach plot in a fertilizer experiment should be protected by guard rows, i.e., by rows that receive the same treatment of the effective plot unit, but are not utilized for observations and harvest. As this technique increases the field work there is a tendency for simplifying it by using, for instance, only one unfertilized row between the adjacent effective plot units. The author studied the viability of the last technique in two fertilizer trials with corn and verified that the yields of the separating (unfertilized) rows increased as the doses of nitrogen applied to the adjacent effective plots were increased, what seems to indicate that these plots were deprived of corresponding quantities of the nutrient in question. In another experiment with corn, in which the established technique was used, the yields of the guard rows were larger than those of the effective plot units, whereas the response to nitrogen was much smaller in the former than in the latter. From lhe results of these experiments the author concluded that at least when studying the effect of nitrogen the simplified teclunique should not be used.