In this paper, we apply a recently developed small-area estimation technique to derive geographically detailed estimates of consumption-based poverty and inequality in rural Shaanxi, China. We also investigate whether using environmental variables derived mainly from satellite remote sensing improves upon traditional approaches that only use household survey and census data. According to our results, ignoring environmental variables in statistical analyses that predict small-area poverty rates leads to targeting errors. In other words, using environmental variables both helps more accurately identify poor areas (so they should be able to receive more transfers of poor area funds) and identify non-poor areas (which would allow policy makers to reduce poverty funds in these better off areas and redirect them to poor areas). Using area-based targeting may be an efficient way to reach the poor since many counties and townships in rural Shaanxi have low levels of inequality, even though, on average, there is more within-group than between-group inequality. Using information on locations that are, in fact, receiving poverty assistance, our analysis also produces evidence that official poverty policy in Shaanxi targets particular areas which in reality are no poorer than other areas that do not get targeted.