This paper is an empirical examination of how a community's income growth is affected by polluting manufacturing activity. The hypothesis I test is that this activity has two conflicting effects: first, industrial investment encourages economic growth through the creation of employment and other positive economic spillover effects and, second, the associated pollution causes out-migration of residents. I hypothesize that a community that is initially relatively wealthy will experience relatively more out-migration of its higher income residents, who are assumed to have a lower tolerance for pollution. Thus, such communities will grow less in response to such investment compared to its poorer neighbors. Therefore, in my econometric model the marginal effect of pollution on income growth is allowed to vary with initial incomes. I use a unique data set that incorporates Toxics Release Inventory (TRI) and census tract-level data for New England for the years 1980 and 1990. The estimated effect of pollution on growth is negative, on average, and is more negative in initial incomes. In an effort to measure the out-migration effects of pollution, I use a separate measure of toxic pollution. I find that, holding constant total pollution, 'toxic' pollution has a more negative effect on growth for wealthier communities. These results are consistent with the above hypotheses.