Double- or two-phase point sampling with a regression estimator is a common forest inventory procedure. When the first-phase sample consists of basal area points only, a diameter distribution based on the first-phase sample cannot be estimated. Thus there is no basis for adjusting the volume distribution of the second-phase sample by regression or ratio estimators within each diameter class. A weighted constrained least squares procedure to align the second-phase stand and stock tables with the first-phase volume and basal area may be applied to obtain a suitably adjusted size class volume distribution. The least squares method described in this paper produces an adjusted stand and stock table with total basal area equal to the first-phase sample total basal area and total volume equal to the regression volume estimate. Because adjustment magnitude can be made inversely proportional to class variance, this method provides better results than adjusting classes by the ratio of second- to first-phase basal area estimates. For. Sci. 37(6):1605-1613.