A two-field experiments was conducted at Tag El- Ezz Agricultural Research Station to evaluate the fungicides Cabrio Top and Amistar, applied as foliar sprays, for control of powdery mildew on six flax cultivars(Giza 9, Giza 10,Sakha 3, Sakha 4,Istro and Jetika) with varying levels of susceptibility to the disease. Disease severity (DS) was used as criterion for evaluating the performance of the tested fungicides. In the first season, the differences in DS between fungicides and the control was not the same for each cultivar, that is, cultivars responded differently to the application of fungicides. Amistar was more efficient than Cabrio Top in reducing DS on all the tested cultivars except Jetika where they showed almost the same level of efficiency. In the second season, the fungicides showed almost the same performance of the first season. Data were statistically analyses according to Smith (1938) and Lessman and Atkins (1963) to estimate optimum plot size, number of replications, and number of sampling units for assessment of powdery mildew severity. Results indicated that increasing plot size decreased the variance per basic unit, the index of soil variability ranged from 0.387 to 0.432. The optimum plot size in Smith,s method was 2.560 m2 and in maximum curvature method was 1.895 m2. The standard error for the severity decreased as number of sample units increased. The optimum number of replications was 4-6 and the number of samples was 8-12 .The required number of samples units to achieve the same level of precision with number of replications indicated increasing precision. When the area of the experimental site is limited, it is better to increase the number of sampling units. The results of the present study suggest that the optimum plot size for assessment of powdery mildew on flax should range from 1.895 to 2.560 m2 as this will improve the chance of homogeneous distribution of natural inoculums (conidia) on plants.