The ability of a degraded soil to respond to successive additions of a toxic organic waste (olive-mill solid waste) and its vermicompost was studied in a controlled incubation experiment for 32 weeks. Hydrolytic enzyme activities (phosphatase, beta-glucosidase), oxidoreductase activities (dehydrogenase, o-diphenol oxidase) and indole acetic acid production, were used as measures of soil perturbation. No microbial activity, indicated by the total lack of dehydrogenase activity, was detected when the olive-mill solid waste was added to the soil. However, after 16 weeks, the activity returned to the original soil levels (1.35 mug INTF g(-1) h(-1)). The addition of vermicomposted olive-mill solid waste increased the original soil dehydrogenase activity by five-fold, indicating a loss of toxicity of the waste during the vermicomposting process; the activity remained high throughout the experiment. At week 21, a second addition of olive solid waste, was made to both olive waste and vermicompost-amended soils, when the soil originally amended with olive waste had reached the activity measured in soil amended with vermicompost. Dehydrogenase activity recovered immediately, reaching levels up to seven-fold higher than the background levels of the soil. The ability of soil to respond to a toxic waste clearly differed after a period of exposure to the waste. The faster response was probably related to the increased pool of stabilized organic matter present in soil, arising from the stabilization of added olive waste in the soil or through the amended vermicompost. The amplitude (period of recovery to the initial state after disturbance) and the elasticity (speed of recovery) of the soil could also be monitored by o-diphenol oxidase and beta-glucosidase activities. However, indole-3-acetic acid production proved to be a useful measure of perturbation only following the second addition of the olive waste. (C) 2004 Elsevier Ltd. All rights reserved.