A study was carried in a rural district in Malawi among HIV-positive individuals placed on antiretroviral treatment (ART) in order to verify if community support influences ART outcomes. Standardized ART outcomes in areas of the district with and without community support were compared. Between April 2003 (when ART was started) and December 2004 a total of 1634 individuals had been placed on ART. Eight hundred and ninety-five (55%) individuals were offered community support, while 739 received no such support. For all patients placed on ART with and without community support, those who were alive and continuing ART were 96 and 76%, respectively (P<0.001); death was 3.5 and 15.5% (P<0.001); loss to follow-up was 0.1 and 5.2% (P<0.001); and stopped ART was 0.8 and 3.3% (P<0.001). The relative risks (with 95% CI) for alive and on ART [1.26 (1.21-1.32)], death [0.22 (0.15-0.33)], loss to follow-up [0.02 (0-0.12)] and stopped ART [0.23 (0.08-0.54)] were all significantly better in those offered community support (P<0.001). Community support is associated with a considerably lower death rate and better overall ART outcomes. The community might be an unrecognized and largely 'unexploited resource' that could play an important contributory role in countries desperately trying to scale up ART with limited resources.