The objective of the present research was to evaluate the use of several spectral vegetation indices (SVIs), including NDVI, SAVI, SR and RSR, obtained from Landsat 7 images, as potential predictors of forest productivity of radiata pine stands. We aimed to evaluate relationships between the variations in stand volume and SVIs over time and the effect of early weed control on stand growth response. We evaluated a large-scale silviculture experiment located at the Central Valley of Chile, since its establishment until 12 years of age, where weed control showed to be the major silvicultural response. Forest inventory measurements were made annually and local equations were used to estimate stand volume. Significant and highly significant correlation was found among SVI ' s and stand productivity parameters. The best relationship was found between NDVI and stand cumulative volume (R-adj=0.92, p-value < 0.0001, RMSE= 0.03), but SR and RSR were able to better track productivity and the major weed control effect on stand volume growth over time. SVIs' coefficient of variation estimates were correlated with estimates of stand productivity variability but no significant relationships were established to provide an index of stand quality due to the sensor spatial resolution and plot sizes. SVIs may serve as important tools to monitor forest growth and high-resolution imagery may provide valuable estimates of stand variability for inventory assessment or as a support tool for growth and yield models.