Corrosion inhibitors are commonly employed to internally protect carbon steel pipelines in aqueous carbon dioxide (CO2)-saturated environments, such as those encountered in oil and gas production and geothermal operation. However, unexpected events can occur that lead to periods where corrosion inhibitor addition ceases completely, or the quantity of inhibitor added falls short of the typical concentration required for the desired level of corrosion mitigation. In these instances of interruption, there is a limited understanding of the inhibitor surface residence time, or ‘persistency’, and the associated effects on the corrosion rate of carbon steel. This study examines the influence of the substrate surface condition and inhibitor concentration on the persistency of benzyldimethyltetradecylammonium chloride (BAC-C14) corrosion inhibitor in a CO2-saturated 1 wt.% NaCl brine at 30°C. An electrochemical rotating cylinder electrode (operating at 1000 rpm) coupled with a dilution process was used to simulate persistency. Experimental results using carbon steel at 0.75x and 1x of the surfactant critical micelle concentration (CMC) showed that inhibitor efficiency and persistency improved markedly at the higher concentration. Supplementary experiments using pure iron illustrated an even stronger interaction between inhibitor and substrate, resulting in BAC-C14 failing to desorb after three days of exposure to uninhibited brine. A first-order kinetic model was assessed in its ability to predict the desorption response after dilution, based on fitting to the inhibitor adsorption response. Whilst a strong agreement was obtained between the theoretical desorption profile and experimental desorption data at 0.75xCMC on carbon steel, the model failed to predict the responses at CMC, as well as those on the pure iron substrate, necessitating consideration of other models.