In gold mining regions, the risk of soil pollution by mercury is a major environmental hazard, especially in tropical areas where soil microflora plays a major part in soil functioning, major bio-geochemical cycles and carbon turn-over. The impact of mercury pollution on soil microflora should thus be carefully assessed in such environments while taking into consideration the specificities of tropical soils. The aim of this study was to compare the effects of mercury (0, 1 and 20 mg of inorganic mercury per gram of soil) on the functional diversity and genetic structure of microbial communities in a tropical soil. We investigated the effects of mercury on tropical soil microflora using soil microcosms spiked with mercury and incubated at 28 8C for 1 month. Microcosm flora, its biomass and its activity, as well as its functional and genetic structure, were followed by cultural methods, measures of respiration, ECOLOG plates, and DGGE (denaturing gel gradient electrophoresis), respectively. Fate of total and bioavailable mercury was estimated by CVAFS (cold vapor atomic fluorescence spectrometry). Results obtained for the microcosms enriched with only 1 mg g_1 mercury were indistinguishable from controls. Conversely, in the presence of high mercury contents (20 mg g(1), an immediate effect was measured on soil respiration, functional diversity (ECOLOG plates) and genetic structure (DGGE), although no significant effect was observed on plate counts or microbial biomass. In addition, whereas microbial activities (respiration and functional diversity) rapidly regained control values, a lasting effect of the high mercury concentration was observed on the genetic structure of the soil microbial community. These modifications took place during the first week of incubation when total mercury concentration was declining and bioavailable mercury was at its highest. This multiple approach study is one of the first attempts at investigating the effects of mercury on soil microbial communities in tropical soils. Our results demonstrate that in the tropical soil under study, mercury affects the soil microbial communities in a different manner than was previously reported in temperate soils. Furthermore, mercury toxicity on soil microbes may be modulated by typical tropical soil characteristics.