Determining the stem volume of a tree, both a standing or felled tree, has long been a chronic problem for foresters, whereas, the need for an accurate estimate of wood volume is becoming more important. The use of Brereton’s formula in Indonesia, nowadays, was based solely on practical consideration in obtaining the data which comprise diameter of the large and small ends of logs; and logs length. However, the relative accuracy of this formula in comparison with other formulae still needs to be tested. The result of this study by using 499 logs in PT Siak Raya Timber (Riau Province) showed that Centroid and Newton’s formulae have the best performance (seen from bias, accuracy and precision) for estimating the log volume of meranti in Sumatera. Bias of both formulae is around zero and very small, therefore it could be neglected. Huber’s formula tends to underestimate the volume, but it has still better performance than Brereton’s, Smalian’s and Bruce’s formulae. Brereton’s and Smalian’s formulae tend to overestimate the volume, whereas Bruce’s formula tends to underestimate. Deviation of errors for Brereton’s is consistently smaller than Smalian’s, and for Bruce’s is the largest one. But, precision for Brereton’s, Smalian’s and Bruce’s formulae are nearly the same.