Theoretical and empirical studies have long suggested that stability and complexity are intimately related, but evidence from long-lived systems at large scales is lacking. Stability can either be driven by complex species interactions, or it can be driven by the presence/absence and abundance of a species best able to perform a specific ecosystem function. We use 64 years of stand productivity measures in forest systems composed of four dominant conifer tree species to contrast the effect of species richness and abundance on three stability measures. To perform this contrast, we measured the annual growth increments of > 900 trees in mixed and pure forest stands to test three hypotheses: increased species richness will (1) decrease stand variance, (2) increase stand resistance to drought events, and (3) increase stand resilience to drought events. In each case, the alternate hypothesis was that species richness had no effect, but that species composition and abundance within a stand drove variance, resistance, and resilience. In pure stands, the four species demonstrated significant differences in productivity, and in their resistance and resilience to drought events. The two pine species were the most drought resistant and resilient, whereas mountain hemlock was the least resistant and resilient, and red fir was intermediate. For community measures we found a moderately significant (P = 0.08) increase in the community coefficient of variation and a significant (P = 0.03) increase in resilience with increased species richness, but no significant relationship between species richness and community resistance, though the variance in community resistance to drought decreased with species richness. Community resistance to drought was significantly (P = 0.001) correlated to the relative abundance of lodgepole pine, the most resistant species. We propose that resistance is driven by competition for a single limiting resource, with negative diversity effects. In contrast resilience measures the capacity of communities to partition resources in the absence of a single limiting resource, demonstrating positive diversity effects.