A photoluminescent turn-off probe was evaluated for the detection of Hg(2+) based on graphene quantum dots (GQDs). Different photoluminescent carbon nanoparticles were prepared using the bottom-up approach, with citric acid and with citric acid mixed with some sulfur compound (thiourea, thioacetamide or glutathione). Dispersions of these GQDs were compared in terms of photophysical behavoior in the presence of Hg(2+) . It was observed that the nanoparticles prepared with citric acid and thiourea (GQDs-TU) presented a better linear photoluminescent suppression (analytical response) in the presence of Hg(2+). Experimental conditions were adjusted to improve analytical response and to obtain analytical figures of merit. A study to evaluate potential interferences in the presence of other heavy metal ions and alkaline and alkaline-earth ions, commonly present in natural water samples. The normalized analytical curve (L0/L versus Hg(2+) concentration) was linear (R(2) = 0.998) up to 0.20 micrograms (L-1). The limit of quantification (LOQ) was 10 micrograms L(-1) and the limit of detection (LOD) was 3 micrograms L(-1) for the batch assay. In addition, the proposed strategy for indirect determination of Hg2+ was applied in a flow injection analysis system (FIA). After optimization of the assay, linear range covered a range up to 900 micrograms L(-1) (R(2) = 0.999) with LOD of 24 micrograms L(-1) and LOQ of 80 micrograms L(-1). The recovery in the fortified water sample was 99.4 percent.