<p>Fuel surcharges in U.S. freight rail are governed by published, formula-based schedules that link per-car-mile charges to fuel price indices. These schedules are disclosed in advance and treated as presumptively compliant under Surface Transportation Board (STB) guidelines. The STB does not, however, audit whether surcharges are assessed as posted, and, to our knowledge, no shipper has challenged a railroad for deviating from its own formula. Using aggregated waybill data from 2009 to 2018, we provide the first causal evidence that assessed surcharges vary systematically with local market conditions despite the uniformity of the posted schedule. A network-based instrument isolates plausibly exogenous variation in railroad concentration. We find that a 10% point increase in a railroad’s local market share leads to an 11% increase in the assessed surcharge. Surcharges also increase with shipper distance and volume but at less than the stated proportional rate. These patterns suggest that railroads selectively underapply surcharge schedules in contested markets while charging closer to the posted amount in captive ones. A back-of-the-envelope calculation suggests that strategic deviations from surcharges that would have been assessed under a benchmark of no market power generated $1.3&#xa0;billion in additional revenue for railroads over the sample period and markets.</p>