In an earlier report, a simple algorithm for effect-assessment on secondary poisoning of birds and mammals was presented. This algorithm (MAR = NOEC/BCF) was drawn up by analyzing an aquatic food chain. In the present study it was tested whether this algorithm can be used equally well for effect-assessment on terrestrial food chains. The pathway soil - earthworm - bird/mammal was used as an example for a terrestrial food chain. Literature data on both bioconcentration factors (BCF-soil-worm) and toxicity data (NOEC-birds/mammals) for six compounds (lindane, dieldrin, DDT, PCP, cadmium and mercury) were studied. As in the aquatic food chain, BCFs can be dertermined both with experimental data and QSARs. However, important differences were found between BCFs for this terrestrial pathway and BCFs for the aquatic pathway analyzed in the previous report. BCFs for the terrestrial pathway are frequently < 1 and rarely above 10, though for the aquatic BCFs upto 10-4 were found for the same compounds. BCFs for the terrestrial pathways seem to depend on soil properties rather than on compound-related properties. Therefore, Maximum Acceptable Risk levels (MAR) can only be calculated for defined situations. By calculating MARs for a standard situation and comparing these to MARs for soil organisms, it was concluded that secondary poisoning could be a critical pathway for cadmium and methyl- mercury. For methyl-mercury secondary poisoning in an aquatic food chain (fish-eating birds and mammals) was also a critical pathway. Secondary poisoning of fish-eating birds and mammals is not likely to occur for cadmium at levels in water below the MAR calculated for aquatic organisms. It was concluded that the simple algorithm can only be used for effect-assessment on terrestrial food chains if details on soil parameters are available.