Conservation policy often incentivizes managers of human-impacted areas to create landscape heterogeneity to maximize biodiversity. In rangeland, patchy disturbance regimes create landscape heterogeneity (patch contrast), but outcomes of heterogeneity-based management are rarely tested for a universal response. We analyzed four habitat variables - vegetation structure, plant functional group composition, litter cover, and bare ground - from five experimental rangelands in Oklahoma and Iowa, USA. We tested for response consistency to heterogeneity-based management across and within locations. We calculated effect sizes for each variable to compare patch contrast on pastures managed for heterogeneity (patch burn-grazing) and pastures managed for homogeneity (grazing with homogeneous fire regimes). Effects varied considerably across and within locations. Effects of heterogeneity-based management were positive for all variables at only three of five experimental rangeland locations. No location showed a consistent pattern of positive effect across all four variables, although one location showed no effect for any variable. At another location, we found a positive effect of heterogeneity-based management on litter cover and bare ground, but no effect on vegetation structure and plant functional group composition. We discuss effect variability and how the fire-grazing interaction applies to rangeland management and conservation. Although it is accepted practice to use heterogeneity-based management to increase rangeland habitat diversity, managers should also confirm that evaluation metrics match desired conservation outcomes.