Concrete has established itself as the most widely used structural material. There is hardly any place where human life and concrete structure do not exist together. It's use is seen in wide variety of structures like buildings, bridges, dams, nuclear structures, floating and submerged structures and so on. Hence, in view of safety, serviceability and economy, proper understanding of the behaviour of concrete is imperative in designing these complex structures. Current reinforced concrete codes are based on strength and serviceability concepts. The tensile strength of concrete is totally neglected in the limit state method of analysis. The concrete in tension is assumed to be fully cracked and conservative method of design is adopted. The crack causes a considerable degradation of stiffness of overall structure and gives rise to regions of stress concentration, which are not accounted for, in the present design methods. Besides, it is found that the size of the structural component significantly influences the stress at failure. It has been fairly well established that large specimens fail by catastrophic crack propagation while small specimens tend to fail in a ductile manner with considerable amount of slow crack growth preceding fracture. Initial attempts to understand the cracking of concrete through the principles of fracture mechanics was made in 1960's. It was concluded that the LEFM and small scale yielding fracture mechanics which are developed for metals are inapplicable to concrete structures except for certain limiting situations such as the behaviour at extremely large sizes. The reasons for the inapplicability of LEFM principles to concrete structures are attributed to slow crack growth, formation of nonlinear fracture process zone, and softening behaviour of concrete in tension. Several analytical and numerical models have been proposed to characterize the fracture behaviour of concrete. In the present work a simple numerical method is proposed to analyse the Mode-I fracture behaviour of concrete structures, using finite element method. The stiffness matrices calculated at the beginning of the analysis are used till the end without any modification. For this reason, the method is named as Initial Stiffness Method (ISM). An attempt has also been made to modify the lattice model existing in literature. The contents of the thesis are organised in six chapters. In chapter 1, a brief introduction to basic principles of fracture mechanics theory is presented. This is included mainly for the completeness of the thesis. In chapter 2, a brief review of literature regarding the application of principles of fracture mechanics to concrete structures is presented. The need for the introduction of fracture mechanics to concrete is presented. Early work, applying LEFM principles to concrete structures is discussed. The reasons for the inapplicability of linear elastic fracture mechanics principles to concrete structures are discussed. Necessities for nonlinear fracture mechanics principles are pointed out. Attention is focused on the influence of the factors like slow crack growth, formation of nonlinear fracture process zone and softening behaviour of concrete in tension on the fracture behaviour. Besides a possible use of fracture energy as an alternative fracture criterion for concrete is contemplated. Several analytical and numerical models (assuming concrete as homogeneous continuum), proposed so far to characterize the fracture behaviour of concrete, are presented and discussed in detail. Different heterogeneous models presented so far are also discussed. In chapter 3, a simple numerical method to analyse the fracture of concrete (strain softening material) in Mode-I, using FEM is proposed. The stiffness matrices are generated only once and are used till the end of the analysis. This feature makes the model simple and computationally efficient. A new parameter namely, strain softening parameter α has been introduced. It is found that this strain softening parameter ‘α’ is a structural property. The results obtained from the present method are found to converge with increasing number of elements thus making the method mesh independent, and thus objective. The method was validated by analysing the beams tested and reported by various researchers. The predicted values of maximum load by the present method are found to agree well with the experimental values. Initially, all the beams are analysed using uniform meshes and load-deflection diagrams are plotted. All the beams are again analysed using graded meshes. The load-deflection, load-CMOD diagrams are plotted from the results obtained from the analysis using graded meshes. In chapter 4, the results obtained in chapter 3 are analysed for size effect. Literature regarding size effect of concrete structures has been reviewed. In addition to the size effect on nominal stress at failure which exists in literature, two new parameters namely, post peak slope and softening slope parameter α have been used to confirm the size effect. This does not exist in the literature. In chapter 5, an attempt is made to modify the lattice model existing in literature. This is done with a view to model concrete as a heterogeneous medium, which would be nearer to reality. The softening property of concrete has been incorporated. The model was validated against some of the experimental results existing in literature. The results are found to be encouraging. The results from this model show the post peak softening similar to the experimentally observed ones. The effects of different probabilistic distributions to the properties of mortar on the maximum load of the beam are studied. It is found that normal distribution of properties to mortar gives the best results. A study is made regarding the sensitivity of various properties of mortar on the maximum load of the beam. It is concluded that load carrying capacity of the beam can be increased by using a mortar of higher tensile strength. Finally in chapter 6, general conclusions and suggestions for further investigations are discussed.