Abstract Wildlife management aims to halt and then reverse the decline of threatened species, to sustainably harvest populations, and to control undesirable impacts of some species. We describe a unifying framework of three feasible options for evaluation of wildlife management, including conservation, and discuss their relative strengths of statistical and causal inference. The first option is trends in abundance, which can provide strong evidence a change has occurred (statistical inference) but does not identify the causes. The second option assesses population outcomes relative to management efforts, which provides strong evidence of cause and effect (causal inference) but not the trend. The third option combines the first and second options and therefore provides both statistical and causal inferences in an adaptive framework. We propose that wildlife management needs to explicitly use causal criteria and inference to complement adaptive management. We recommend incorporating these options into management plans.