We examine the implications of workhorse trade models for how aggregate productivity, real GDP and real consumption, as measured by statistical agencies, respond to changes in trade costs. In a range of models, changes in measured productivity are equal to the inverse of an export-share weighted average of changes in variable trade costs incurred domestically. Under certain conditions, despite the multiple biases in the CPI, measured real consumption captures the first-order effects of changes in variable trade costs on welfare. Through the lens of these results, we interpret some of the empirical work on measured gains from trade.There is a large literature that uses structural models to evaluate welfare and productivity gains from reductions in trade costs (see e.g. Costinot and Rodriguez-Clare 2014, Melitz and Redding 2014, and references therein). There is also an extensive empirical literature estimating the link between trade and aggregate measures of economic activity such as real GDP, aggregate productivity, and real consumption (see e.g. Pavcnik 2002, Romer and Frankel 1999, Feyrer 2009, and Broda and Weinstein 2006. The connection between theoretical and measured gains from trade is not immediate. For example, it is well known that the consumer price index (and hence real consumption) does not reflect changes in the number, composition, and quality of available varieties (see e.g. Boskin Comission 1996 and Hausman 2003), which constitute an important source of welfare gains in recent trade models. Similarly, as we discuss below, GDP deflators (and hence aggregate productivity) may not reflect the reallocation of production across heterogeneous producers emphasized in these models. In this paper we ask whether the gains