The frequently anecdotal nature of evidence concerning the impact of warfare on conservation poses numerous problems and there have been calls to apply a strict set of conditions to such data to improve the rigor of scientific analysis in this field. To illustrate the difficulties, however, of applying strict quantitative conditions on such data a deterministic model of conflict-linked deforestation in sub-Saharan Africa was constructed and the implications of the model discussed. Our model indicates that from 1990–2005 approximately 35,000 ha of timber have been used to support officially recorded UN refugees in this region: this is a continuing impact, albeit quantified using data with some potential error. An alternative semi-quantitative approach was also used, with reported environmental impacts of conflict assessed for reliability and severity using a number of empirical criteria. Data focusing on the Democratic Republic of Congo and Rwanda were subsequently analysed using this framework. Illegal resource exploitation was identified as the primary impact resulting from conflict and, in some instances, a driver of the hostilities. From the joint consideration of the conflict and post-conflict phases such exploitation is concluded to be the product of lawlessness and anarchy generated by violent uprisings rather than violence per se. As such, armed conflict does not pose a novel threat to protected areas but rather amplifies threats extant during peace, creating a need for appropriate responses by those involved in conservation management. With both the occurrence and violence of conflicts in sub-Saharan Africa increasing, the impacts of warfare are pertinent to both the immediate and long-term management of biological resources in the region.