Whether free returns create economic value for online retailers is a question of increasing importance. Using order-level data from an online fashion retailer in Sweden, we investigate the introduction of free returns for customers in Denmark starting November 1, 2017, but not for customers in Sweden, Norway, Finland, France, Germany, and the rest of Europe (the control group). Using difference-in-differences, with and without sample matching, the introduction of free returns increased the order bill by 54.95 Swedish Krona (SEK) (9.15%), product variety per order by 0.057 (8.74%), gross margin per order by 29.90 SEK (9.71%), and returns by 0.195 items per order (7.86%). This likelihood of increased returns was generally similar across different product categories. The findings are robust to parallel trends, event studies, placebo effects, alternative control groups (with Scandinavian countries, Sweden and Norway), country-time trends, and correlations between outcome variables. Based on back-of-the-envelope calculations, the cost of offering free returns is greater than the increase in gross profit driven by free returns. Though our results show that free returns may not be of short-term benefit, retailers must also consider the long-term strategic implications of free returns.