The assumptions in the does not effect the response on the neighboring plots and the treatments assocated with plots in a block is constant . In many experiments, especially in agriculture, the response on a given plot may be affected by treatment on neighboring plots as well as by the treatments applied to that plot there fore, the estimates of treatment differences may therefore devated because of interference of neighboring units plots. Neighbor balanced block designs are used for modeling and controling inter ference effects between neighboring plots which all the designs are assumed such as in the direction of the blocks (say left neighbor and right_neighbor effects). Optimality of designs under interference model is studied. In this thesis, three different types of models M 1, M 2 and M 3 in related to neighbor effects are considered when observation are uncorrelated. These models are: When y i;j is the response from the i_th plot in the j_th block, is the effect of the j_th block, is the left_neighbor effect due to the treatment in the (i-1)_th plot of jth block. and are the right neighbor effect due to the treatment in (i+1)_th plot in j_th block. " i;j is error terms independently with mean zero and variance M 1 represents the model with one sided neighbor effect (say left). M 2 and M 3 correspond to the model with nudifferentiated and differentiated left and right neighbor effects, respectively. In this thesis circular neighbor-balanced designs are considered. Also, we present a general approach to determine the optimal designs for contrasts among direct treatment effects that can be useful for many kinds of interference models. We consider an experiment for comparing v treatments in b blocks of size k with the one dimensional arrangement of plots in each block. We used, generalized kushner’s method for finding optimal repeated measurments designs to find optimal designs under an interference model. The universaly optimal block designs with the same parameter. Too, this the method is applied for finding optimal designs when observation independent. By using this method, the optimality of circular neighbor balanced block designs for direct treatment effects and total effects are considered which errors in it are independent. Also, in this thesis, discuss’s optimal incomplete block designs and universally optimal of these designs under M 1 , M 2 and M 3 are investigated.