Fire ecology has a long history of empirical investigation in rangelands. However, the science is inconclusive and incomplete, sparking increasing interest on how to advance the discipline. Here, we introduce a new framework for qualitatively and quantitatively understanding the ranges of variability in fire regimes typical of experimental investigations in rangeland fire science compared with the range of conditions that actually occurred during contemporary social-ecological times. We implement this framework for one of rangelands' most pyrogenic systems—the Great Plains of North America. We identify four social-ecological fire eras that have epitomized people's relationship with wildland fire in the Great Plains since the last glacial maxima. These cultural fire eras include the now-extinct coexistence era (indigenous fire use), the suppression era (extermination of wildland fire occurrence), the shadow era (localized prescribed burning groups), and the emerging wildfire era (resulting from wildland fire management failures, continued decoupling of human-fire ignitions, and global change). Our synthesis demonstrates that experimental fire conditions have not explored the types and ranges of variation in fire regime components responsible for shaping rangeland vegetation—now, in the past, or into the future. Instead, scientific investigations have focused largely on controlling and minimizing sources of uncertainty and experimental variation, essentially eliminating ranges of variation that underpin the functioning of fire in modern social-ecological systems. Yet a series of scientific investigations exist that targeted a wider range of variability in fire regime components, leading to major advancements and the rejection of a number of long-standing rules of thumb in rangeland science and management. These include 1) the manipulation of fire return interval, 2) the pyric herbivory experiments, and 3) the extreme fire trials. We discuss the general philosophy shared among these studies, introduce scientific standards needed to avoid common pitfalls, and highlight opportunities to better understand how rangeland pattern and process correspond to critical ranges of variation in the human-fire relationship.