Quantification of water transient storage (A_s) has become critical in biogeochemical studies of stream ecosystems addressed to examine nutrient retention. It is expected that higher A_s may enhance interaction between nutrients and biota and thus, increase nutrient retention. However, results from literature are controversial. We hypothesize that besides the quantity of A_s, the quality or nature of it can also be a key factor influencing nutrient retention. In this study we tested our hypothesis by performing 16 constant rate additions of phosphate, ammonium, and chloride in 4 reaches of a semi-natural canal where we introduced 3 types of naturally colonized substrata packs (mud, sand and cobbles) to modify A_s. To assess nutrient retention, we estimated nutrient uptake rate (U) and the uptake coefficient and used the OTIS model to estimate A_s. At the end of the experiment, and to understand which type of substrata pack contributed more to nitrogen retention, we performed a ¹⁵NH₄ addition. The introduction of substrata packs increased water residence time, decreased average water velocity and, in the case of “mud” packs, increased A_s. As we hypothesized, nutrient retention differed among treatments being the “cobble” treatment the one with highest U-PO₄ and ¹⁵N signature.