Light Detection and Ranging (LiDAR) remote sensing has been established as one of the most promising tools for large-scale forest monitoring and mapping. Continuous advances in computational techniques, such as machine learning algorithms, have been increasingly improving our capability to model forest attributes accurately and at high spatial and temporal resolution. While there have been previous studies exploring the use of LiDAR and machine learning algorithms for forest inventory modeling, as yet, no studies have demonstrated the combined impact of sample size and different modeling techniques for predicting and mapping stem total volume in industrial Eucalyptus spp. tree plantations. This study aimed to compare the combined effects of parametric and nonparametric modeling methods for estimating volume in Eucalyptus spp. tree plantation using airborne LiDAR data while varying the reference data (sample size). The modeling techniques were compared in terms of root mean square error (RMSE), bias, and R2 with 500 simulations. The best performance was verified for the ordinary least-squares (OLS) method, which was able to provide comparable results to the traditional forest inventory approaches using only 40% (n = 63; ~0.04 plots/ha) of the total field plots, followed by the random forest (RF) algorithm with identical sample size values. This study provides solutions for increasing the industry efficiency in monitoring and managing forest plantation stem volume for the paper and pulp supply chain.