The single largest item in the United States foreign aid health budget is antiretroviral therapy (ART) for the treatment of HIV/AIDS. Many supply- and demand-side factors in sub-Saharan Africa could cause smaller than expected epidemiological effects of this at scale drug provision. We provide what appears to be the first quasi-experimental evidence on the effect of at scale drug provision in a poor country, using the phased roll-out of ART in Zambia, a setting where approximately 1 in 6 adults are HIV positive. Combining anthropometric data from national household surveys and a spatially-based triple difference specification, we find that local ART introduction increased the weight of high HIV likelihood adult women. This finding from a clinically difficult setting suggest that the generalized challenges of scalability of ART for adult health in sub-Saharan Africa are surmountable.