This study aimed to investigate the possibility to remove mercury (Hg) using the live or dead biomass of Hg resistant purple nonsulfur bacteria (PNSB) for their potential application to clean up shrimp ponds and during cultivation. The efficiency of Hg<sup>2+</sup> removal by three PNSB strains; <i>Rhodovulum sulfidophilum</i> SRW1–5, and Afifella <i>marina</i> strains SSS2-1 and SSW15-1, grown with 2 different growth conditions (microaerobic light and aerobic dark conditions) were tested (2 mg/L HgCl<sub>2</sub>, 4.5 mg dry cell weight/mL, 30 min). It was found that the efficiency to remove Hg<sup>2+</sup> by the dead cells of all strains was significantly higher than for live cells; and the most effective strain was SSS2-1. The highest Hg<sup>2+</sup> removal under the optimum conditions using a 4 mg/L initial HgCl<sub>2</sub> concentration by live and dead cells grown with both incubation conditions were approximately 87% and 95%, respectively. Under both incubation conditions, Hg<sup>2+</sup> biosoption by strain SSS2-1 fitted the Freundlich model for live cells and the Langmuir model for dead cells. The kinetics of Hg<sup>2+</sup> biosorption by both cell types of strain SSS2-1 suggests the process as a pseudo-second order kinetic model. This study demonstrates that biomass of the strain SSS2-1 has great potential for its biosorption to remove Hg<sup>2+</sup> from contaminated areas like shrimp ponds.