Ruptures and other injuries to tendons and ligaments are among the most common damages to the body, specifically in the young and physically active population. Recently, tissue engineering (TE) has been recognized as a new approach to regenerate damaged tissues instead of commonly used surgical grafts such as biological and synthetic grafts. On the other hands, some shortcomings of tendons and ligaments restoration may be overcome through tissue engineering methods. The aim of this research is to model and design the silk/P3HB hybrid yarn scaffold for tendon and ligament tissue engineering. First of all, a neural network method used to estimate mechanical properties of multilayer twisted scaffold. Then, a genetic algorithm model was developed for automation and optimization designing scaffold and afterward a computer model of finite elements method was used for simulation. So, variable parameters including the number of filament and the number of twist in each layer of 4-layer twisted scaffold were defined and according to Taguchi orthogonal matrix, the experiments were designed. After that, the designed neural network was trained based on the back propagation algorithm. The best neural network was formed from one hidden layers, in which the hidden layer has 8, 3 neurons for output layer, 8 neurons for input layer. Tangent-hyperbolic, sigmoid and linear was used as activation functions. In following step, genetic algorithm method was applied to optimize the main characteristics successfully with 10% error in comparison with real values of tissues. Finally, a hybrid scaffold made of silk fibroin and P3HB electrospun fibers was designed. It was necessary to optimize the electrospinning parameters for P3HB solutions. Optimized conditions of electrospinning were found by using statistical analysis. The concentration, voltage, distance spinneret-collector and solution flow rate conditions were optimized as 5%, 10 kv, 20 cm and 0.5 ml/h for P3HB solution. Then the hybrid scaffolds were produced and physico-mechanical properties as well as in vitro cell culture were experienced. The results of FTIR tests showed that there is no any interaction between silk fibroin and P3HB electrospun fiber in structural formation of hybrid scaffold where both basic polymers remain in the final scaffold, independently. Also, both structures including silk fibroin and hybrid samples made of P3HB and silk were manufactured to evaluate the growth of L929 cells while no toxicity was proved in all samples according to in vitro cytotoxicity. The results of cytocompatibility tests show that the samples combined with P3HB electrospun fibers have more cell growth than SF sample in longer time (3 rd day) and they can be a promised substrate for L929 cells. Key words: Tendon and ligament tissue engineering, Artificial neural network, Genetic algorithm, Finite element method, Hybrid yarn