Although pp'DDT usage was strongly limited or banned in most parts of the world during the last three or four decades, the parent compound, its homologues and their metabolites still occur at levels which might pose a risk for many ecosystem components. A case of DDT pollution of industrial origin was discovered in 1996 in Lake Maggiore, the second largest (212 km2) and deepest (370 m) lake in Italy, causing concern for wildlife and human health. The extensive monitoring of many biotic and abiotic compartments which followed from 1998 in order to assess the pollution level and its trend in time, provided a great availability of data referring to DDT contamination of the different fish species of the lake. In this study, the recent contamination levels in selected fish species were compared to those measured in 1998 to evaluate the temporal pollution trend of the lake and its natural recovery, given that no remediation measures were carried out on the contaminated soils and sediments in this time span. Moreover, a modelling approach to test the equilibrium condition between water and pelagic fish species was used. Analytical results of pp'DDT and pp'DDE concentrations in lake water were used as input data in the bioenergetic model by Connolly & Pedersen (1988) to calculate concentrations in two fish species and to compare the predicted and the measured contamination. Sampling and analytical determination of DDT homologues in lake water: Five water sampling campaigns were carried out from May 2002 to February 2004 in three sampling sites of Lake Maggiore. Suspended and dissolved pollutants were determined separately. Quantitative DDT homologue analyses were performed by HRGC coupled with ECD detection by the external standard method. Single water extracts were put together in correspondence with the stratification zones of the water column inferred on the basis of the temperature profile to improve analytical sensitivity. Selection of fish data: Concentrations of DDT and DDE in fishes were selected from recent literature (CIPAIS 2003, 2004). Bioaccumulation model: The bioenergetic model proposed by Connolly & Pedersen (1988) was used to assess the bioaccumulation of pp'DDT and pp'DDE of Alosa fallax (landlocked shad) and Coregonus spp. (whitefish), selected among the different species as representative of a secondary consumer level. The average concentrations of pp'DDT and pp'DDE in water to be used as input data in the bioenergetic model were obtained considering all the concentrations measured at the three sampling stations in the epylimnion where the fish species considered in this study spend most of their life. The resulting values were 0.05 and 0.16 ng/L for pp'DDT and pp'DDE, respectively. Average measured pp'DDT and pp'DDE concentrations in landlocked shad were 0.81 +/- 0.39 and 1.69 +/- 0.71 mg/kg lipids, respectively, and were 0.29 +/- 0.12 and 1.06 +/- 0.41 mg/kg lipids for the whitefish. Calculated and measured values turned out to be in quite good agreement for pp'DDT, while measured pp'DDE concentrations were higher than expected on the basis of the bioenergetic model in both species. Probably metabolic transformations of pp'DDT accumulated in fish tissues in the past are responsible for the observed differences between calculated and expected pp'DDE concentrations in fish. Pelagic fishes of Lake Maggiore seem to maintain the DDT accumulated during their life time and the most efficient mechanism responsible for the fish population recoveries is probably their generation changes; for this reason, equilibrium models cannot be used until negligible pp'DDT concentrations are reached in fish tissues. The limit proposed for pp'DDT in water by the EU Directive 2000/60, which will come in force in 2008, is 0.2 ng/L, four times higher than the average concentration measured in Lake Maggiore waters. Nevertheless, concentrations measured in Lake Maggiore fish were very close and sometimes exceeded the Maximum residue limits (MRLs) settled by the Italian legislation for foods (0.1 mg/kg w.w. for fish containing 5-20% lipid). It seems, therefore, that the 'environmental quality standard' of 0.2 ng/L cannot guarantee the suitability of fish for human consumption.