Predictions and recent observations of changing frequencies and intensities of climate extremes have prompted ecologists to increasingly study their ecological impacts. Rising interest in this field of research reflects growing recognition that changing climatic variability can impact ecological dynamics independent of climatic means, and that the ecological impacts of climate extremes may be of equal or greater magnitude than gradual changes in mean climate. However, recent concerns have emerged that traditional approaches used to understand and quantify relationships between climate and ecological processes may not be predictive of responses to extreme climatic conditions with no historic analog. In this dissertation, I describe tests of current knowledge about how precipitation impacts ecosystem processes by considering how changing extremity at both intra-annual and interannual timescales impacts different components of the carbon cycle. To achieve this, I employed a novel experimental design that imposed multiple levels (n = 11 levels, n = 4 replicates), and thus a gradient, of precipitation amount and extremity within a single growing season. These manipulations were imposed within two intact ecosystems of opposing climatic backgrounds; the semi-arid steppe of Colorado (low mean productivity) and the mesic tallgrass prairie of northeastern Kansas (high mean productivity). I show that despite these ecosystems harboring differing ecological characteristics, aboveground net primary productivity was consistently more sensitive to extreme wet years than severe drought, and thus carbon gains during wet years were greater than drought-induced productivity reductions. Despite asymmetrical productivity responses to precipitation extremes in both systems, there was consistent evidence for an underlying linear relationship as best describing the response of productivity to changes in growing season precipitation within these grasslands, in agreement with current models. Coupling this experimental data with long-term records within the mesic grassland revealed strong interactions between variability in rainfall patterns within and among years. Variability in intra-annual rainfall patterns, and in particular large and more variable event sizes, acted to magnify the reductions in ecosystem functioning during drought. A systemic review of the literature adds further complexity to these dynamics from an organizational perspective, suggesting that both the response and recovery of ecosystems to climate extremes are mediated by ecological responses and interactions that propagate from the individual, population, to the community-level to collectively impact ecosystem-level functioning. Overall, my research demonstrates a critical role for changes in precipitation extremity at both intra and interannual timescales and levels of ecological organization with respect to predicting the dynamics of ecosystem functioning amid climate change.