We describe change in weight, midupper arm circumference (MUAC), arm muscle area (AMA) and arm fat area (AFA) in 1130 pregnant HIV-infected women with CD4 counts > 200 as part of the BAN Study ( www.thebanstudy.org ), a randomized, controlled clinical trial to evaluate antiretroviral and nutrition interventions to reduce mother-to-child transmission of HIV during breast feeding. In a longitudinal analysis, we found a linear increase in weight with a mean rate of weight gain of 0.27 kgs/week, from baseline (12 to 30 weeks gestation) until the last follow-up visit (32-38 weeks). Analysis of weight gain showed that 17.1% of the intervals between visits resulted in a weight loss. In unadjusted models, MUAC and AMA increased and AFA declined during late pregnancy. Based on multivariable regression analysis, exposure to the famine season resulted in larger losses in AMA [-0.08, 95% CI -0.14, -0.02; p = 0.01] while AFA losses occurred irrespective of season [-0.55, 95%: -0.95, -0.14, p = 0.01]. CD4 was associated with AFA [0.21, 95% CI 0.01, 0.41, p = .04]. Age was positively associated with MUAC and AMA. Wealth was positively associated with MUAC, AFA, and weight. While patterns of anthropometric measures among HIV-infected, pregnant women were found to be similar to those reported for uninfected women in sub-Saharan Africa, effects of the famine season among undernourished, Malawian women are of concern. Strategies to optimize nutrition during pregnancy for these women appear warranted.