A new method for using sample tree heights and diameters to calibrate (localize) height/diameter and volume/diameter equations for a given stand is presented and illustrated with Scots pine data. The height/diameter and volume/diameter equations are assumed to contain random parameters that vary from stand to stand. The means, variances and both within-equation and between-equation covariances of the random parameters are estimated using data where volumes of trees from several stands are accurately measured. In applications, the random parameters of the height and volume equations of a new stand are predicted from sample tree heights and diameters using the covariance structure of parameters and linear prediction theory. When the number of sample trees was small, the new method predicted volumes of tallied trees better than a traditional method. For. Sci. 37(3):781-801.