Based on data trees inventoried in 10-years period on permanent samplings plots (n= 1,604) in three forest management units (Grčarice, Logatec and Vzhodne Haloze) we analysed quality of data, fellings of growing stock, and assess their accuracy using different algorithms. We compared calculated fellings of growing stock to the official harvest statistic. Five year increment of cutting trees in ten years period should be taken into account when calculating fellings of growing stock since the latter value would be underestimated for approximately 10 %. Confidence interval of felling is similar when using algorithm for (simple) random sampling or algorithm for systematic sampling. Fellings of growing stock per sampling plots were not normally distributed; however, comparison to the results using Bootstrap algorithm indicated, that procedures based on normal distributions are allowable.