The present work aims to describe the nitrogen (N) budget in integrated aquaculture systems with Nile tilapia (Oreochromis niloticus) and Amazon River prawn (Macrobrachium amazonicum) in earthen ponds, with and without the addition of different substrates. The experimental design was completely randomized, with three treatments (without a substrate, with a geotextile fabric substrate, and with a bamboo substrate) and four replications. Diet was the major input of N in the systems, ranging from~65 to 71% and followed by inlet water (~26-31%). The portion retained in reared animals and periphyton ranged from~21 to 25% (being~21-24% in fish and prawns). The outputs that contributed most to the accumulation and release of N were, respectively, sediment (~24-38%) and N2 (~30-36%) emitted to the atmosphere. The addition of substrates did not improve the accumulation of nitrogen in the biomass of the target species. This suggests that the periphyton had a minor role on feed availability. In general, the systems were not efficient in using nitrogen since only~22% of all available nitrogen was retained into prawn and tilapia biomass. On the other hand, the emission of N 2 (an inert gas) to the atmosphere almost compensated the nitrogen supplied in the diet that was not assimilated by the reared animals and periphyton. In addition, data suggest that the integrated aquaculture in stagnant ponds may sequester substantial amounts of nitrogen from nutrient-rich aquatic environments and could be used as a mitigation tool.