A uniformity test was utilized in two field trials each included 720 units (one basic unit = 0.3 m²). The trials were conducted at Side experi­mental research station in 1998/99 and 1999/200 seasons using the lentil variety Giza 51. The objective of this study was to determine the optimum plot size and shape. In analysis using Smith's method, the soil heterogeneity index was 0.7622 and 0.8424 in the first and the second seasons, respectively. The optimum plot size estimated by Smith's meth­od was ranged from 0.9m² to 1.5m². Results of analyses using the mod­ified maximum curvature technique indicated a plot size range was 2.1­2.4 rn². While, the results of analyses using the comparable variance (V) and relative information estimate (RI) referred a plot size of 1.5m². Plot shape has no significant effect on plot-to-plot variability. Since these methods are based on different criteria, it is expected that the esti­mates of plot size may not agree with each other. Thus we recommend that the optimum plot size in lentil (net harvested plot area) should be 0.09m² with increasing number of replications.