In constructed wetlands (CWs), bed substrate and leaf leachates from vegetation may correct the low C/N ratio that constrains heterotrophic denitrification and nitrate removal from irrigated agricultural drainage water. However, the interactive effects of bed substrate type and leaf leachates on denitrification are still unknown. By focusing on a CWs pilot plant, we designed a laboratory experiment to evaluate i) wether denitrification potential rates varied among bed substrates: calcareous gravel (a conventional substrate), gravel+soil from a natural wetland (silty loam Solonchak, 1.5 % of organic C) and gravel+biochar from pyrolyzed ornamental plants (75 % of organic C); and ii) the response of denitrification within each bed substrate to the addition of their respective leaf leachates. We found that denitrification potential rates were lower in gravel beds (0.011 ± 0.006 μgN2O-N gDM−1 h−1) than those observed with the addition of biochar (0.06 ± 0.03) and especially soil (0.78 ± 0.04), with soil being the most advantageous option. Besides, leaf leachates addition boosted denitrification rates in all cases. Nevertheless, the effect of leachates was relatively higher in gravel beds than in the other substrates (15 times higher vs. 2 and 4 times with soil and biochar, respectively). Our outcomes highlight limited denitrification when using gravel substrate not only by low C but also due to essential macro- and micro-elements, and support the role of plant leaves as internal and self-sustainable source of nutrients.