Initial attack (IA) success has long been one of the primary performance measures used by agencies with wildland firefighting responsibility in the United States (US) and elsewhere. The US federal agencies currently state that (1) they credit an IA success when an ‘unwanted’ wildfire is uppressed before it expands beyond 100 acres of forest or 300 acres of grass or brush, and (2) the US Forest Service and Department of Interior strive to achieve 98 percent and 95 percent initial attack success rates, respectively. Achievement of these levels is often hailed as extraordinary success, but, as we explain here, environmental factors alone (e.g. weather, fuels, terrain) will tend to constrain the majority of wildfires to < 300 acres, regardless of suppression activities. Thus, a size-based IA success metric is a poor proxy for actual firefighting effectiveness. Moreover, none of the agencies’ fire-reporting or decision-support systems have ever used the terms ‘wanted’ or ‘unwanted’ to classify individual fires, thereby engendering significant ambiguity surrounding perceptions of what effectiveness and success look like. Furthermore, an emphasis on high IA success rates may be counterproductive from long-term ecological and fire-management perspectives. The challenge is to develop alternative performance measures that are less ambiguous and that better align with risk management principles. We discuss risk-based performance measurement from the perspective of linking decisions to actions to outcomes, and offer recommendations for broad-scale, consistent metrics that can be aggregated at meaningful scales. A key insight is the necessity of upstream assessment and planning to both guide and establish an evaluative framework for downstream fire management decisions. Current federal policy recognizes that both fire control and fire use have a place in the fire-management paradigm, but fire use has been limited largely due to entrenched disincentives. Strategic objectives (e.g. protection, restoration, maintenance) set forth in spatial land- and fire-management plans and associated performance measures can provide the context within which the necessary incentive structure can develop.