The South Nation River basin in eastern Ontario, Canada is characterized by mixed agriculture. Over 1600 water samples were collected on a bi-weekly basis from up to 24 discrete sampling sites on river tributaries of varying stream order within the river basin between 2004 and 2006. Water samples were analyzed for: densities of indicator bacteria (Escherichia coli, Clostridium perfringens, enterococci, total and fecal coliforms), the presence of pathogenic bacteria (Listeria monocytogenes, E. coli O157:H7, Salmonella spp., Campylobacter spp.), and densities of parasite Giardia cysts and Cryptosporidium oocysts. Relationships between indicator bacteria, pathogens, and parasite oocysts/cysts were overall weak, seasonally dependent, site specific, but primarily positive. However, L. monocytogenes was inversely related with indicator bacteria densities. Campylobacter, Salmonella, Giardia cysts and Cryptosporidium oocysts were most frequently detected in the fall. E. coli O157:H7 was detected at a very low frequency. Exploratory decision tree analyses found overall that E. coli densities were the most utilitarian classifiers of parasite/pathogen presence and absence, followed closely by fecal coliforms, and to a lesser extent enterococci and total coliforms. Indicator bacteria densities that classified pathogen presence and absence groupings, were all below 100 CFU per 100 mL(-1). Microorganism relationships with rainfall indices and tributary discharge variables were globally weak to modest, and generally inconsistent among season, site and microorganism. But, overall rainfall and discharge were primarily positively associated with indicator bacteria densities and pathogen detection. Instances where a pathogen was detected in the absence of a detectable bacterial indicator were extremely infrequent; thus, the fecal indicators were conservative surrogates for a variety of pathogenic microorganisms in this agricultural setting. The results from this study indicate that no one indicator or simple hydrological index is entirely suitable for all environmental systems and pathogens/parasites, even within a common geographic setting. These results place more firmly into context that robust prediction and/or indicator utility will require a more firm understanding of microorganism distribution in the landscape, the nature of host sources, and transport/environmental fate affinities among pathogens and indicators.