The study provides empirical evidence on how minimum tracking error varies, as the stringency of a sustainability criterion is varied. The sustainability criterion is based on environmental (EV) ratings for a universe of large capitalization U.S. firms. Increasingly sustainable portfolios are created from increasingly smaller subsets each containing stocks with increasingly higher EV ratings. Minimized tracking error standard deviation increases with sustainability stringency and varies from 0.4% per year for a portfolio, created from the 400 stocks with the highest EV ratings to 4.6% per year for a portfolio, created from the 20 stocks with the highest EV ratings. These sustainable portfolios’ tracking errors appear to be equal or lower than those of existing sustainable funds from similar universes.