Abstract: Of the five variables studied, viz, latewood percent, diameter at breast height, age of wood, radial annual growth rate and relative position of breast height of sample point, the latewood percent was found to be the most important factor, with the help of which, breast height density of wood of sample trees from natural stand of masson pine can be estimated with reasonable accuracy. In multiple regression analysis in the presence of latewood percent, estimative efficience don’t increase with the number of variables contained. The common measural characteristics can’t be used to estimate the density of wood at breast height.