Transmissible pathogens in source-separated human urine, intended for reuse in agriculture, mainly originate from faeces that cross-contaminate the urine. The health risks associated with the enteric pathogens will largely be dependent on their initial concentration and their inactivation during storage in the urine. Faecal sterols have proven stable in urine and can, rather than indicator bacteria, be used to quantify the faecal cross-contamination. In this study, urine collection tanks were sampled and ratios between various faecal sterols were used to determine if the urine was contaminated by faeces. Twenty-two percent of samples from the upper part of the tanks and 37% of samples from the bottom sludge were found to be contaminated. Coprostanol concentrations in the contaminated urine samples corresponded to a mean faecal contamination of 9.1+/-5.6 mg l(-1) urine. E. coli was absent in a majority of the samples. Faecal streptococci and clostridia were enumerated but not found to correlate with coprostanol concentrations in contaminated samples.