The authors are of the opinion that computational models suggested for exterior beam column connections in literature are not practicable in application as most of them are too complex to be used for hand analysis. Practicable computational models on the other hand contradict with the experimental data. In this paper, a simpler strut and tie model with a new truss model will be used to predict the failure modes and loads of the tests of an experimental database consisting of monotonically loaded exterior beam column connections. The relative contributions of the truss and strut mechanisms to joint shear will be determined and some design recommendations will be made. It is apparent from the results that the simple model predicts the failure loads and modes as accurate as other complicated models. The parametric studies carried out when developing the model shows that stirrups contribute as high as 29.7% to joint shear strength if they are provided by a stirrup ratio of at least 0.0055. It should also be noted that not all the stirrups yield within this region. In joints with stirrup ratios of 0.003 to 0.0055, all the stirrups yield and the contribution to joint shear strength is 18.14%. The stirrups of ratios less than 0.003 in the joint do not contribute to the joint shear strength of joints. The new simple strut and tie model proposed predicts the failure loads and modes very accurately. The model is applied on a database of 74 tests and the average of the ratio of the predicted failure load to the actual failure load is 0.99 for all tests. In this paper, several parametric studies will be carried out on this simple model both with and without stirrups in order to figure out this simple model's reliability. The influence of parameters such as column axial load, column bar ratio, concrete cylinder strength and the joint aspect ratio on the predicted joint shear strength of the model will be compared with the results of the authors' parametric studies on experimental data.