Introduction The oil-shale basin of Estonia (the Kohtla- Järve region off the north-western coast of Estonia) is the biggest industrially used oil- shale basin in the world. Estonian oil-shale is used to fuel power stations (90% of Estonian electricity is produced in thermal power stations) and as a raw material in the petrochemical industry. To date, approximately 200 million tonnes of oil- shale solid waste (ash) from thermal power stations has been deposited on an area of about 2000ha near Narva, Estonia. In addi- tion, 80 million tonnes of solid waste (ash formed in the retorting process of oil-shale) from the Kiviter petrochemical factory has been deposited in large spent shale piles (ash heaps) near Kiviõli and Kohtla-Järve. These ash heaps cover approximately 250ha and a further 1.4 million tonnes of ash are added annually. The sulphur-rich phenolic wastewater (3000–8000m 3 /day) that is leaching from the ash heaps (ash- heap water [AHW]) is not purified in the local biopurification facilities due to its (supposed) high toxicity to the activated sludge micro-organisms (1–3). Until the Spring of 1997, AHW was dis- charged into the River Kohtla and River ATLA 27, 359–366, 1999 Joint SSCT/ETS Congress 359 Predicting the Toxicity of Oil-shale Industry Wastewater by its Phenolic Composition Anne Kahru, Lee Põllumaa, Rain Reiman and Annely Rätsep Institute of Chemical Physics and Biophysics, Akadeemia tee 23, Tallinn 12618, Estonia Summary — The chemical composition and toxicity of five phenolic wastewater samples col- lected from the Kohtla-Järve (Estonia) oil-shale industry region were analysed. The total phenolic contents (HPLC data) of these samples ranged from 0.7mg/l to 195mg/l. A total of 11 phenolic compounds were found in the wastewater samples, the most abundant being phenol (up to 84mg/l) and p-cresol (up to 74mg/l). Artificial phenolic mixtures were also com- posed, to mimic the content of phenolic compounds in the wastewater samples. The theoret- ical toxicities of these artificial mixtures were calculated by using the toxicities of the individual phenolic constituents to photobacteria (the BioTox TM test) and were assumed to have an additive mode of action. From the BioTox data, the additive toxic effects of pheno- lic compounds in the artificial mixtures were confirmed to be highly probable. The toxicities of the wastewater samples and their artificial phenolic analogues (mixtures) were studied by using a battery of Toxkit microbiotests (Daphtoxkit F TM magna, Thamnotoxkit F TM , Pro- toxkit F TM and Rotoxkit F TM ) and three photobacterial tests (Microtox TM , BioTox TM and Vibrio fischeri 1500). The wastewaters were classified as toxic (two samples), very toxic (two samples) and extremely toxic (one sample). Comparison of the test battery responses showed that the industrial wastewaters were 2–28-fold more toxic than the respective artificial phe- nolic mixtures. The photobacterial tests proved to be the most appropriate for screening pur- poses. This was the first attempt to use a test battery approach in the toxicity testing of Estonian wastewaters. The study showed that the toxicity of oil-shale industry wastewaters could not be predicted solely on the basis of their phenolic composition, since only 7–50% of their toxicity was shown to be due to phenolic compounds. It is true, to a certain extent, that the majority of environmental samples are usually very complex and contain various types of pollutants. As even a full chemical analysis (which is very expensive) can easily miss the constituent(s) with the greatest toxic effect(s), the use of toxicity tests in parallel to chemi- cal analysis should be encouraged. Key words: microbiotests, photobacteria, phenolic wastewaters, mixtures, toxicity testing.