Abstract: In this study, face recognition in angled status using invasive weed optimization algorithm is studied. Invasive weed optimization algorithm is one of the most recent optimization algorithms in evolutionary algorithms family. Combined with other existing algorithms, invasive weed optimization algorithm proposes an appropriate approach in order to an improved solution. In this study, a combination of fuzzy cost function and invasive weed optimization algorithm is used in order to human face recognition. Fuzzy logic is used in order to face recognition in angled status and invasive weed algorithm is used in order to obtain the optimal threshold to find the favourable solution. Tests conducted on the proposed algorithm shows 91.4% of existing faces in MIT database.