Currently, studies investigating the carbon balance in forest ecosystems are particularly relevant due to the global increase in CO2 content in the atmosphere. Due to natural reforestation over the past 25–30 years, birch (Betula pendula Roth.) forests were extensively grown and established on abandoned agricultural lands in Bashkir Cis-Ural (Republic of Bashkortostan, Russia). The significant positive aspect of reforestation on fallow lands is the carbon sequestration that takes place in the tree phytomass, especially at the growth stage of stand formation. The aim of this article is to test the approach of using a UAV-mounted LiDAR camera to estimate the phytomass and carbon stocks in different-aged birch forests growing on abandoned arable lands in Bashkir Cis-Ural. The methodology was developed using 28 sample plots, where the LiDAR survey was performed using a DJI Matrice 300 RTK UAV. Simultaneously, the stand characteristics and phytomass of stem wood were also estimated, using traditional methods in the field of forest science. The regression equations of phytomass dependence on stand characteristics at different stages of reforestation were constructed using data obtained from LiDAR imagery. It was shown that the above-ground tree biomass could be precisely estimated using the index obtained by multiplying the number of trees and their average height. A comparison of the data obtained using traditional and LiDAR survey methods found that the accuracy of the latter increased in conjunction with stand density. The accuracy of estimation ranged from 0.2 to 6.8% in birch forests aged 20 years and over. To calculate carbon stocks of the above-ground tree stands, the use of regional conversion coefficients is suggested, which could also be applied for the estimation of carbon content in trunk wood and leaves. An equation for the calculation of above-ground biomass carbon stocks of birch forests on abandoned arable lands is proposed.