A study to determine an adequate plot size for rice trials was conducted in Bagaces, Guanacaste, Costa Rica. A uniformity trial was planted, and the data collected were used to apply the maximum curvature method. The point of maximum curvature -determined by visual inspection- where the plotted curve for the relationship between plot size and variance coefficient stops falling drastically and becomes constant, was located between 15 m<span name="style_sup">2</span>  and 20 m<span name="style_sup">2</span> .The point of maximum curvature was estimated, using F-tests between the mean variances of each plot size, to be 16m . Nevertheless, for the sake of reliability, the adequate plot size was established at 20m<span name="style_sup">2</span>. Using the method that Fairfield Smith proposed in 1938, an index of soil heterogeneity value of 0,83 was obtained, indicating that the soil can be classified as heterogeneous (this measurement's scale goes from 0 to 1, where 0 indicates perfect homogeneity and 1 indicates perfect heterogeneity).. Using the method that Fairfield Smith proposed in 1938, an index of soil heterogeneity value of 0,83 was obtained, indicating that the soil can be classified as heterogeneous (this measurement's scale goes from 0 to 1, where 0 indicates perfect homogeneity and 1 indicates perfect heterogeneity).

plot size; rice; maximum curvature method; soil heterogeneity; Smith's index