Estimation of under-canopy radiation is crucial for characterizing vegetation–energy interactions and for a better understanding of its implications for ecosystem studies and forestry applications. Under-canopy radiation regimes are difficult to model due to the complex interaction of light with vegetation structure. Also, measuring radiation under the canopy over large areas is challenging using traditional field-based procedures. In this context, LiDAR remote sensing shows great potential for radiation estimation because it directly measures the three-dimensional canopy structure. The primary aim of this dissertation is to improve the understanding of under-canopy light regime using discrete return LiDAR and estimate solar radiation in forests with different structural characteristics. Based on the availability of LiDAR data, research sites were chosen in the coniferous forests of Sierra National Forest (SNF), California, and a chronosequence of mixed deciduous forest plots located in the Smithsonian Environmental Research Center (SERC), Maryland. First, LiDAR-derived digital surface models with and without vegetation canopy were used to assess the first-order effect of vegetation on solar radiation in SNF. The results showed a significant difference (p value < 0.001) in insolation values between the two surface models, with the mean solar irradiation over the bare surface almost three times higher than vegetation canopy surface. Next, a ray-tracing method was used to estimate beam radiation using LiDAR point clouds, and estimates were compared with in situ pyranometer measurements across three forest plots in SERC and were found to be in good agreement (RMSE = 13.94 W/m2). Lastly, LiDAR-derived vertical light transmittance values were compared with measurements from field-based PAR sensors, across five forest plots in SERC and were found to be in good agreement (R2 = 0.84). These results suggest that LiDAR remote sensing can provide reliable fine-scale estimates of beam radiation and vertical transmittance values under the vegetation canopy without the need for extensive ground measurements. This information provides a better understanding of radiation variability under the canopy and can help potentially improve the estimates from a range of land surface models such as snowmelt and hydrological models, and possibly help downscale general circulation model (GCM) predictions.