Rape is known to require more sulfur (S) than is necessary for the secondary metabolite synthesis. It is hypothesized that its coarse root system harbors more homogenous microbial biomass specific for sulfur immobilization=remobilization. In a growth chamber experiment, this work examined the ability of rape and barley to take up the labile 35 S (including 35 S-SO 4 2À and 35 S-organic extracted by hot water). For that, the endogenous compounds of soil organic matter were previously labeled with 35 S-SO 4 2À for three months prior to plant growth. The use of ''rhizobag'', a polyamide bag, easily allows the separation of rhizosphere from non-rhizosphere soil. In order to simulate the high input agrosystems, the soil received just before sowing, a unique dose of 76.9 mg N kg À 1 soil as ammonium nitrate and three levels of S (20.4, 30.8, and 61.5 mg kg À 1 soil) as MgSO 4 . Despite the dilutions made by levels of additional S-SO 4 2À , the results showed an increase in 35 S uptake by rape in contrary to barley, which showed a progressive decrease with increasing S-SO 4 2À dilutions. The mean percentage values of total 35 S taken up by rape were higher than barley (8.9, 22.2, and 28.2% at day 20, 42, and 56 after sowing vs. 3.7, 17.1, and 21.8% respectively). Correlated to 35 S-uptake, significant coefficient was found with the 35 S-SO 4 2À (0.78, p < 0.01), and 35 S extracted by hot water at 70 C for 18 h in rape rhizosphere (r 2 ¼ 0.75, p < 0.05) but not in barley. Similarly, significant correlation coefficient was observed between soil arylsulfatase and root 35 S uptake of rape (r 2 ¼ 0.79, p < 0.01) but not of barley. These results clearly showed the higher capacity of rape rhizosphere in regulating via the arylsulfatase activity the uptake of 35 S-endogenous compounds.