Forest health is a serious problem worldwide. Knowledge of the intensity and causes of rapidly declining forests is important for managing various forest types under the scenario of climate change. In the present study, we report on the health conditions of four young conifer forests by applying dendrochronological techniques to measure unexpected growth decline in Xinzhou District, northern Shanxi Plateau. A total of 86 samples in the form of increment cores were obtained from the study sites. The widths of tree-rings from all the samples were measured and cross-dated, then each individual tree series was standardized to identify health conditions using spatial characteristics of temporal growth. We defined unhealthy tree growth intervals as being a decrease in the growth of the individual (tree-ring index < 0.9) for at least five years when the mean tree-ring index was less than 0.7 and at least one year of tree-ring index values reached a minimum of 0.5. We calculated the percentage of trees that were unhealthy in each year, and defined an unhealthy event for the site if greater than 30% of the trees were in decline. We found three major unexpected decline growth events in the majority of individual trees in the four forests, these being 1910-1940, 1970-1987, and 1990-2012. However, the growth declines in each forest showed deviations in intensity and duration such as declines were more vigorous in the KeLan forest during the 1930s than in the other forests. To better understand the causes of the unhealthy events, a correlation analysis was performed for the three tree ring width chronologies and the meteorological records. The correlation coefficients showed that forest growth was negatively associated with May-June temperatures. However, precipitation revealed positive correlation in July and the Palmer Drought Severity Index showed positive correlation over the growing season of May-September. To validate these results, we compared our study with historical records and local paleoclimate studies. The drought events recorded in the historical documents and other paleoclimate reconstructions reasonably matched our results. These results suggest that extreme drought events caused by persistent high temperature or reduction of precipitation in the growing season are the main driving factors for radial growth decreases in trees, which was the main reason for the spatial coherence of unhealthy events. The differentiation of local stand characteristics or tree species might play a role in the inconsistency of the time characteristics. Therefore, the present study provides useful information for forest health, which has positive and realistic significance for assessing the impact of extreme climatic conditions on the healthy growth of forests and for developing reasonable forest protection measures.