Frequently, mechanical components are subjected to complex loads, in which its directions vary with time, possibly leading to a failure by multiaxial fatigue. In these conditions, designing by uniaxial fatigue criterium might provide unconservative life predictions. Usually, standards and design books don’t mention much of new comes at multiaxial fatigue. On the other hand, innumerous multiaxial fatigue failure criterium can be found in literature. This dissertation, has a main goal of verifying some of these criteria, in which critical plane approach is used, such as: Findley, McDiarmid, Goodman Generalized, Brown-Miller, Fatemi-Socie, Smith-Watson-Topper e Liu. For that, it was developed an algorithm capable of computing multiaxial fatigue life by critical plane method. The algorithm includes a double layer amplitude filter: by partition and racetrack filter; two cycle counting methods: Critical plane Rainflow and Modified Wang-Brown; a path equivalent method by Moment of Inertia; as well as all multiaxial fatigue models listed above. The behavior of an 304L Steel subjected to traction x torsion history loads was verified with the algorithm. It was concluded that, for this situation, Fatemi-Socie and Brown-Miller models showed better predictions.