Culture-independent fecal source tracking methods have many potential advantages over library-dependent, isolate-culture methods, but they have been subjected to limited testing. The purpose of this study was to compare culture-independent, library-independent methods of fecal source tracking. Five laboratories analysed identical sets of aqueous samples that contained one or more of the following sources: sewage, human feces, dog feces, cattle feces and gull feces. Two investigators used methods based on PCR amplification of Bacteroidetes marker genes and both successfully discriminated between samples that did or did not contain human fecal material. One of these investigators was also able to identify the remaining sources, except for gull, with a low rate of false positives. A method based on E. coli toxin genes successfully identified samples containing sewage and cattle feces, but missed some samples with human feces because of low marker prevalence in individual human fecal samples. Researchers who used community terminal restriction fragment length polymorphism (T-RFLP) were limited by the amount of DNA recovered from samples, but they correctly identified human and cattle fecal contamination when sufficient DNA was obtained. Culture independent methods show considerable promise; further research is needed to develop markers for additional fecal sources and to understand the correlation of these source-tracking indicators to measures of human and environmental health.