A solution containing 35SO42−and 3H2O was applied to four plots (5 × 5 m) in a boreal coniferous forest in the Laflamme Lake watershed, Québec, under two contrasting conditions: in summer (plots 1 and 2), and on the snowpack before snowmelt (plots 3 and 4). The transit of both these tracers in the soil solution was then followed through a network of soil lysimeters located at different depths. Four months after the summer application, 3H2O had infiltrated the whole soil profile at plot 1, while 35SO42−was only observed in the LFH and Bhf horizons. A 35SO42−budget calculated from mid-August to November indicated that 89 and 10.6% of the added 35SO42−was retained within the LFH and the Bhf layers, respectively. Fifteen months later, the added 35SO42−was distributed in the following proportions within the soil horizons: LFH (73.7%), Bhf (11.8%) and Bf (12.8%), for a total retention rate of 98.3%. The superficial penetration of 3H2O at plot 2 was indicative of a major lateral water movement that prevented the calculation of a 35SO42−budget. This situation also was observed at plot 4 during snowmelt. At plot 3, 3H2O moved freely through the soil profile and a significant fraction of the added 35SO42−reached the B horizons, where it was presumably adsorbed on aluminum (Al) and ferric (Fe) oxides. The 35SO42−budget for plot 3 from March to November indicated that 87% of the added 35SO42−was retained within the soil profile, with most being retained in the B horizons (LFH = 33.1%, Bhf = 33.1%, Bf = 20.8%). The contrasting retention patterns of 35SO42−within the soil profile following the summer addition and snowmelt likely was caused by the contrasting soil temperatures and soil solution residence times within the different soil layers. The persistence of 35SO42−in the soil solution of the entire profile long after the initial tracer infiltration, and the relative temporal stability of specific activity of SO42−, point to the establishment of an isotopic equilibrium between the added 35SO4 and the active S-containing reservoirs within a given soil horizon. Overall, the results clearly illustrate the very strong potential for 35SO42−retention and recycling in forest soils.