The thesis addresses the Job Shop Scheduling Problem (JSSP), considering the processing time as a variable that follows a probability distribution approaching a productive environment, which leads to catalog it as a Stochastic Job Shop Scheduling Problem where the objective function is set as the minimization of the time in which all jobs are finished (makespan). The SJSSP is treated as a combinatorial optimization problem of NP-Hard type due to its complexity, therefore, simulated annealing is used as a solution method that emulates the heat treatment of steel where the temperature of a certain state conditions the decision making, all this to escape from local optima. The algorithm starts the search for feasible solutions by storing them and selecting (at the end of the execution) the best schedule found, i.e. with the lowest makespan. During the paper the process is shown from the treatment of the processing time, algorithm development, calibration, generation and representation of schedules, analysis of obtained results, conclusions and future work. Finally, the proposed algorithm has shown a good makespan performance when compared with recognized instances of the Job Shop Scheduling Problem addressed by several authors.