This study considers long-term dynamics of the vegetation indices derived from Landsat imagery of the period from 1989 to 2014. The mass death and damage of forests by Siberian moth occurred in dark coniferous taiga in 1994–1996. The images have covered both altered (disturbed) and background forest conditions. The case study of the forest massive is located in the model area of the Angara Taiga Region of Eastern Siberia (Krasnoyarsk krai). The values of spectral brightness of satellite images are represented by produced index images of NDVI vegetation index and shortwave vegetation index SWVI. The study employed forest-surveying materials from 1992 and data on the degrees of forest damage during the first (in 1995) and second (in 1996) year of Siberian moth outbreak. The latter was obtained by forest-pathology research using production-scale spectrozonal aerophotography. The average background values of indices have gradually decreased with the forest age or remained constant, excluding the young growth stage and 200-year-old light coniferous forests. High seasonal variability of NDVI compared to SWVI is the key to use satellite data for the timeframe of a week in June for every year to analyze long-term dynamics. We have approved the conclusion of other researchers that both the mean and the coefficient of variation of SWVI are the most informative in the evaluation of the degree of Siberian moth damage of forests. We recommend recognizing three degrees of forest damage by Siberian moth—light, moderate, and severe (continuous). Long-term dynamics of vegetation indices in disturbed forests has been marked with demutation and inhomogeneity of stand cover. Taking into account the intensity and frequency of wildfires, it is highly probable that reforestation will not occur in dark coniferous forests. Therefore, the regular satellite monitoring of the sites of Silk moth occurrence would be useful.