The spatial and temporal distributions of indicator bacteria in a small, multiple-use source drinking water reservoir in Southern California, USA were quantified over the period August 2001-July 2002. High levels of total and fecal coliform bacteria were present in Canyon Lake (annual geometric mean concentrations+/-SEM of 3.93+/-0.02 and 3.02+/-0.03 log cfu/100mL, respectively), while comparatively low levels of enterococci and E. coli were found (1.16+/-0.02 log cfu/100mL and 0.30+/-0.03 log MPN/100mL, respectively). As a result, these different indicator bacteria yielded quite divergent indices of water quality, with 72.1% of all surface samples (n=294) exceeding the USEPA single-sample limit of 400 cfu/100mL fecal coliform bacteria, while none (0%) of the samples exceeded the single-sample limit for E. coli (n=194). Regression analyses found a positive correlation between total and fecal coliform bacteria (R=0.50, significant at p<0.001) and between enterococci and E. coli (R=0.51, significant at p<0.001), but no correlation or inverse correlations were found between coliform concentrations and enterococci and E. coli levels. External sources responsible for the high total and fecal coliform bacteria were not identified, although laboratory studies demonstrated growth of the coliform bacteria in lake water samples. Enterococci and E. coli were not observed to grow, however. Bacteria concentrations varied relatively little laterally across the lake, although strong vertical gradients in fecal coliform and enterococcus bacteria concentrations were present during summer stratification, with concentrations about 10x higher above the thermocline when compared with surface concentrations. In contrast, total bacteria, total virus and total coliform bacteria levels were unchanged with depth. Seasonal trends in bacteria concentrations were also present. This study shows that the choice of indicator bacteria and sampling depth can both strongly affect the apparent microbial water quality of a lake or reservoir.