A Monte Carlo experiment is carried out on collected data to compare the precision of volume‐ and volume increment estimates. The volume calculation is based on stratification according to diameter. All diameters are known, but only a fraction of the heights. Five different procedures for stratified sampling are compared. To compare the procedures a modified empirical coefficient of variation is calculated. Given sample fraction, there is no difference in CV neither for the volume estimate nor the volume increment estimate between the five sampling procedures. The CV for the volume estimate is varying from 1.7 to 2.0% for 10% sample fraciton, while doubling the fraction gives a CV from 1.2 to 1.5%. Considering the volume increment estimate there is a benefit including sample trees from the previous registration when sampling at the next. When previous sample trees are not retained a CV from 13.2 to 15.4% for 10% sample fraction and from 9.6 to 11.0% for 20% are obtained. The corresponding when previous sample trees are retained are 7.1 to 8.1 % and 4.8 to 5.9%.