To estimate the plot size in experiments with cassava, an experiment, using the cultivar "Cramuquém" was conducted at the Campus of the "Universidade Estadual do Sudoeste da Bahia" in Vitória da Conquista, Brazil. The maximum curvature method, the modified method for maximum curvature, the method for comparison among variances were used. The estimate of the plot size changed with the applied methodology and the analyzed characteristic. The modified method for maximum curvature was the one which allowed for the obtainment of more adequate estimates. By applying this methodology and considering that the ideal plot would make possible an efficient evaluation of all analyzed characteristics, the adequate size of the plot was found to be 26.59 m² (44 plants).
Manihot esculenta Crantz; coefficient of variation; method for maximum curvature; experimentation
