Forest co-management programs aim to conserve forest resources. However, there is little evidence of its effectiveness. We assess the impact of co-management approaches on forest conditions in Zomba-Malosa and Ntchisi forest reserves in Malawi using a multiple-site, plot-based, and control-intervention design. We used tree density and species richness as indicators of forest condition. Evidence of human activities was used as potential indicators of current and future impacts. Local peoples’ perceptions of the impact of co-management on forest were also sought to validate the inventory information. Co-managed plots have higher tree density than state managed plots. Indicators of human activities including felled trees, farming and settlement plots, and grazing, were observed in both co-managed and state managed forest block. A majority of respondents, 84% in Zomba-Malosa and 73% in Ntchisi, perceive the co-management program to have a positive impact on forest conditions against a general worsening trend. Despite having a potential to improve forest conditions, the findings suggest that the outcomes of a co-management may vary depending on preexisting conditions and how communities understand and interpret the program. Hence, programs should not be implemented as a universal package. Furthermore, even with method triangulation, the lack of baseline data limited the quantification impacts; hence, integration of participatory research into the program is recommended.