Increasing evidence suggests that the use of a single bioassay will never provide a full picture of the quality of the environment. Only a test battery, composed of bioassays of different animal and plant species from different trophic levels will reduce uncertainty, allowing an accurate assessment of the quality of the environment. In the present study, a test battery composed of 20 bioassays of varying biological endpoints has been compared. Apart from lethality and reproductive failure in earthworms, springtails, nematoda, algae and vascular plants, these endpoints also included bioavailibility of metals (bacteria), heat-shock induction (nematodes, algae), DNA damage (bacteria, earthworm, vascular plants), beta-galactosidase (Daphnia) and esterase activity (algae) and a range of immunological parameters (earthworm). Four chemicals (cadmium, phenol, pentachlorophenol and trifluralin)--each representing a different toxic mode of action--were applied in a dilution series (from 1 mg/kg up to 1000 mg/kg) onto OECD standard soil. The tests have been performed both on these artificially contaminated soil samples and on aqueous extracts subsequently obtained from these soils. The results show that the immunological parameters and the loss of weight in the earthworms were among the most sensitive solid-phase assays. Esterase inhibition and heat-shock induction in algae were shown to be extremely sensitive when applied to soil extracts. As previously shown at the species level, no single biological endpoint was shown to be the most sensitive for all four modes of toxic action.