Pollution-induced community tolerance (PICT) was used to study effects of phenol on soil bacteria. Phenol was added to an agricultural soil in a microcosm experiment. The effects were studied for up to four months. Bacterial growth rates were estimated with the leucine incorporation technique. This technique was also used as detection method for PICT. Changes in community structure were studied using the phospholipid fatty acid (PLFA) pattern. Increased phenol PICT of the bacterial community was found at phenol concentrations above 1 μmol/g wet weight soil. Direct inhibiting effect on bacterial growth rates 1 d after adding phenol was correlated to PICT. Phenol toxicity was reflected by changes in the structure of the bacterial community, although PICT appeared more sensitive than the PLFA method. In soil amended with 1 to 10 μmol phenol/g soil, bacterial growth recovered within one week. In the soil amended with the highest phenol concentration (30 μmol/g soil), bacterial growth rate recovered from total inhibition after 27 d, eventually reaching values six times higher than in the control. However, PICT did not change during the four months the experiment was performed. The specificity of PICT was also studied by examining cotolerance to 2-chlorophenol, 2,4-dichlorophenol, 2,3,6-trichlorophenol, Cu, and Zn. Adding phenol induced cotolerance of the bacterial community to the other phenols, although always at a lover level than to phenol. No cotolerance was found to metals in phenol-polluted soil. We conclude that the PICT concept is a valuable tool in determining phenol toxicity to bacterial communities, especially in situations where bacterial growth has recovered. Cotolerance between different phenols can, however, make interpretations of PICT more complicated.