This paper employs multivariate regression to forecast the calorie intake of Bangladeshi farm households in the hunger season based on the household income, production, and demographic composition in the current (post harvest) season. Nutritional vulnerability profiles are derived from the estimation of ex ante mean and variance of future consumption. The results show the income increase induced by introducing transgenic rice will reduce each individual household's probability of suffering a future consumption shortfall and its vulnerability. The overall vulnerability profile of farm households improves in Bangladesh.