In this work we analyzed the sterol content of agricultural soils. Three eukaryotic sterols, cholesterol, beta-sitosterol, and ergosterol were chosen as representative of the animal, plant, and fungal kingdoms, while coprostanol was validated as a marker of human fecal matter contamination. Three soils subjected to different treatments (sewage sludge application, irrigation by saline waters, and contamination by industrial and municipal wastes) were sampled and their sterol content was measured and compared with adjacent untreated soils. The effects of time, location, and treatment were evaluated by means of a number of statistical techniques. Beta-sitosterol concentration varied from 0.9 to 30 mg kg(-1). Lesser values were measured in Cremona (2.1 mg kg(-1)) than in Bari (4.0 mg kg(-1)) and Naples (10.9 mg kg(-1)) soils. No significant effects were detected for cholesterol and ergosterol. Coprostanol was present after sewage sludge disposal and contamination by industrial and municipal wastes, while it was absent in the soil treated with saline water and in the adjacent untreated soil. Coprostanol concentration did not vary much within site and time of sampling, with a mean value of 0.2 mg kg(-1). We confirmed coprostanol as a useful persistent marker of human fecal matter contamination. Multivariate analysis highlighted a clear distinction between the eukaryotic sterols and coprostanol. In addition, a different behavior between ergosterol and cholesterol on one side and beta-sitosterol on the other was detected. This preliminary work suggests that sterols deserve a deeper study of their use as indicators in agricultural soils.