In the present work, a new polycyclic fatigue damage model is formulated and applied for short glass fibre reinforced thermoplastics. The model is able to capture experimental trends observed for the considered composites. The damage growth description involves a set of 20 parameters in the case of a complete 3D –structure. In the current paper, it is considered the particular case of a displacement controlled fatigue tensile test involving 4 damage parameters. The present contribution is a first approach of parameter identification. It is considered a least squares sense based cost function and homogeneous fatigue tests performed on a short glass fibre reinforced polyamide. The identified set of parameters appears to be not depending on the adopted initial values. The model as the parameters determined by the minimisation algorithm, are validated on a fatigue test performed with a different loading condition.