When the aim of an experiment is the estimation of a Generalised Linear Model (GLM) , standard designs from linear model theory may prove inadequate A lso many experiments measure a response that cannot be adequately described by a linear model with normally distributed errors . In this experiments , the response variables has a non-normal distribution and the exponential family . Also , such a response variables impressible from one or several explanatory variables that may be discrete or continuous . On the other hand , sometimes, they may be correlated and are often run in blocks of homogeneous experimental units . In order to find such a design , the proper design should be planed . This thesis describes an approach for finding design of experiments to estimate GLMs through the use of D-optimality. Furthermore , we will discuss on the suitable statistical model and design for non-normal and correlated data , and then the estimation of the parameters will be found by the Generalized Estimating Equations (GEE). Finally , For constructing such a design , the algorithm will be implemented . Then, robust D- optimal exact block design based on equal or unequal blocks size will be discussed and by using of simulation, we will show that the resulting design is acceptable .