To obtain robust data on the toxicity of LAS, tests with the collembolan Folsomia candida L., the oligochaetes Aporrectodea caliginosa Savigny (earthworm) and Enchytraeus crypticus Westheide and Graefe (enchytraeid) were performed in a sandy loam soil. Additionally limited tests with LAS spiked to sewage sludge, and subsequently mixed into soil, were performed. For the endpoint of interest, reproduction in soil, we found an EC10 of 205 mg LAS kg(-1) soil [8.6-401] [95% confidence limits] for F. candida and an EC10 of 46 mg LAS kg(-1) soil [13-80] for A. caliginosa after 28 days. E. crypticus was not affected by concentrations up to 120 mg LAS kg(-1) soil. When adding (low contaminated) non-spiked sludge to soil, high stimulation of reproduction was observed for E. crypticus and A. caliginosa but not for F. candida. We argue that this difference in stimulative response between the tested species is related to the difference in feeding behaviour. Sludge spiked with LAS did not significantly affect the reproduction of F. candida (fertility: number of juvenile offspring) and A. caliginosa (fecundity: number of cocoons) (dose equivalent to 181 g and 91 g LAS kg(-1) sludge, respectively). Significantly reduced reproduction was observed for E. crypticus (at 120 mg LAS kg(-1) soil+sludge corresponding to 72 g LAS kg(-1) sludge) compared to non-spiked sludge. The reproduction by E. crypticus was, however, comparable to the reproduction observed in the control soil without sludge. Compared to LAS directly spiked to soil, the reproductive output of organisms exposed to spiked sludge was either not significantly different (F. candida, E. crypticus) or significantly improved (A. caliginosa). More studies are needed in order to make firm conclusions on the potential effect of artificially contaminated sludge in soil systems.