Ocean acidification (OA) progression is affected by multiple factors, such as ocean warming, biological production, and runoff. Here we used an ocean-biogeochemical model to assess the impact of river runoff and climate variability on the spatiotemporal patterns of OA in the Gulf of Mexico (GoM) during 1981-2014. The model showed the expected pH and aragonite saturation state (ΩAr) decline, due to the increase in anthropogenic carbon, with trends close to values reported for the Subtropical North Atlantic. However, significant departures from the basin-averaged pattern were obtained in part of the northern GoM shelf, where pH and ΩAr increased. Model sensitivity analyses showed that OA progression was counteracted by enhanced alkalinity from the Mississippi-Atchafalaya River System. Our findings highlight that river alkalinity is a key driver of carbon system variability in river-dominated ocean margins and emphasize the need to quantify riverine chemistry to properly assess acidification in coastal waters.