Two ways for the selection over the pierced stems on account of the cane borer (Diatraea saccharalis (Fab.)) are tried, along with eight sampling methods through which the placement of samples within the experimental area is combined into different posicioning, ranging from border sides -this method regularly practiced in such estimations- to the division of the experiemental area into quadrants. The results arrived at in methods that solely observe corners and diagonals show estimations of the average which range over the results turned out by simple randoming method and quadrants method. Allowing for the high degree of variability of clusters of Diatraea sacharalis (Fab.) in a given area, we state with finality that the estimation worked out through the method of the quadrants is more reliable and the selection of sugar cane stems should be carried out disregarding subjective criteria. Damaged stems show better characteristics than nuckles.