Agro-industrial wastes are used as carbon precursors in the production of activated carbon because they are rich in lignocellulosic materials. This study aimed to investigate the relationship between the main characteristics of precursor materials and activated carbons by multivariate analysis. After characterization of the precursor materials and their respective carbons, the principal component analysis and canonical correlation analysis were performed. Materials with cellulose/lignin ratio > 3.0 led to the production of carbons with a higher pore diameter, while materials with a cellulose/lignin ratio <= 1.0 led to a low pore volume. Using the mathematical models obtained, it is possible to predict the carbon characteristics using only the composition data of the lignocellulosic materials.