Floodplain restoration is popular to address excess nutrients, but its ability to enhance photolysis of emerging contaminants has not been evaluated. We used the numerical model MIKE-21 to simulate photolysis reactions within the inundated surface water of restored floodplains along a mid-size river. We examined both "high" and "low" floodplain scenarios where inundation occurs 5% (storms) and 50% (baseflow) of the year, respectively. We simulated photolysis of the pharmaceuticals morphine, codeine, and methamphetamine and, for context, compared it with nitrate removal (denitrification and plant uptake). Pollutant removal due to floodplain restoration was greater for the low floodplain (e.g., 18.8% for morphine) than for the high floodplain (5.6% for morphine) due to greater water exchange relative to channel flow. The fastest- and slowest-reacting pollutants (morphine and methamphetamine, respectively) were always transport- and reaction/kinetics-limited within floodplain surface water, respectively. Yet, those with intermediate decay-rate constants switched from reaction limitation to transport limitation as the floodplain length increased, and removal leveled off at an optimum length of ∼1000 m. However, as the floodplain width increased, the required floodplain length for 30% removal decreased. Optimal restored floodplain conditions for photolysis would maximize light exposure, which may differ from those for nutrients.