Rapid population growth is often cited as an important correlate of high poverty rates in lowincome countries. As a result, much thought and many resources have been put into designing policies which address both poverty and the "population problem". We investigate the conceptual and empirical bases for these views and policy responses, drawing on a recent household-level survey from Bangladesh, one of the poorest and most densely populated countries in the world. We find that allowing for even modest returns to scale in household consumption reverses the oft-cited positive association between low income and large household size. Thus, adding children to a household in our sample is likely to be much less costly than often thought, and the deleterious consequences for poverty may be considerably overstated. The most pressing issues instead appear at the level of communities and of individuals within households. At the level of communities, available evidence suggests that pollution, congestion, and environmental degradation form a coherent basis for relating population growth to poverty; however, specific evidence for these forces remains scant. At the level of individuals, there is considerable evidence that poverty is associated with the relatively poor treatment of women and girls in Bangladesh, and this is exacerbated by high population growth. First, levels of maternal mortality are high, so the risk of death during childbirth is non-trivial. On average, between 2 and 3 women out of every 100 will die in the process of childbirth. Second, son-preference in fertility patterns is pronounced; roughly 60% .of youngest children are boys versus 50% for other children. This suggests that in many families, fertility rates are being pushed upward by son-preference. Third, rates of excess female mortality are high (as many as 8% of women are "missing"), and the rates appear to be positively associated with high overall fertility rates. Thus, high fertility is both a product and a source of gender inequality. The evidence suggests that we should shift the focus in the poverty and population growth debate to consider impacts on individuals within households.