This work proposes a computational algorithm based on a spectral selection NIR range for soil organic carbon percentage estimation. The results show that the proposed algorithm outperforms traditional methods by up to 0.5 R 2 .