The effect of solar UV-A irradiance and solar UV-A dose on the inactivation of Escherichia coli K-12 using solar disinfection (SODIS) was studied. E. coli K-12 was seeded in natural well-water contained in borosilicate glass tubes and exposed to sunlight at different irradiances and doses of solar UV radiation. In addition, E. coli K-12 was also inoculated into poly(ethylene) terephthalate (PET) bottles and in a continuous flow system (10 L min(-1)) to determine the effect of an interrupted and uninterrupted solar dose on inactivation. Results showed that inactivation from approximately 10(6) CFU mL(-1) to below the detection level (4 CFU/mL) for E. coli K-12, is a function of the total uninterrupted dose delivered to the bacteria and that the minimum dose should be >108 kJ m(-2) for the conditions described (spectral range of 0.295-0.385 microm). For complete inactivation to below the limit of detection, this dose needs to be received regardless of the incident solar UV intensity and needs to be delivered in a continuous and uninterrupted manner. This is illustrated by a continuous flow system in which bacteria were not fully inactivated (residual viable concentration approximately 10(2) CFU/mL) even after 5 h of exposure to strong sunlight and a cumulative dose of >108 kJ m(-2). This has serious implications for attempts to scale-up solar disinfection through the use of re-circulatory continuous flow reactors.