Amplified fragment length polymorphism (AFLP) analysis was used to investigate the genetic diversity in isolates of the ectomycorrhizal fungus Cenococcum geophilum from serpentine and non-serpentine soils in Portugal. A high degree of genetic diversity was found among C. geophilum isolates; AFLP fingerprints showed that all the isolates were genetically distinct. We also assessed the in vitro Ni sensitivity in three serpentine isolates and one non-serpentine isolate. Only the non-serpentine isolate was significantly affected by the addition of Ni to the growth medium. At 30 microg g(-1) Ni, radial growth rate and biomass accumulation decreased to 73.3 and 71.6% of control, respectively, a highly significant inhibitory effect. Nickel at this concentration had no significant inhibitory effect on serpentine isolates, and so the fitness of serpentine isolates, as evaluated by radial growth rate and biomass yield, is likely unaffected by Ni in the field. In all isolates, the Ni concentration in the mycelia increased with increasing Ni concentration in the growth medium, but two profiles of Ni accumulation were identified. One serpentine isolate showed a linear trend of Ni accumulation. At the highest Ni exposure, the concentration of Ni in the mycelium of this isolate was in the hyperaccumulation range for Ni as defined for higher plants. In the remaining isolates, Ni accumulation was less pronounced and seems to approach a plateau at 30 microg g(-1) Ni. Because two profiles of Ni accumulation emerged among our Ni-insensitive serpentine isolates, this result suggests that different Ni detoxification pathways may be operating. The non-serpentine isolate whose growth was significantly affected by Ni was separated from the other isolates in the genetic analysis, suggesting a genetic basis for the Ni-sensitivity trait. This hypothesis is further supported by the fact that all isolates were maintained on medium without added Ni to avoid carry-over effects. However, because AFLP analysis failed to distinguish between serpentine and non-serpentine isolates, we cannot conclude that Ni insensitivity among our serpentine isolates is due to evolutionary adaptation. Screening a larger number of isolates, from different geographical origins and environments, should clarify the relationships between genetic diversity, morphology, and physiology in this important species.