The microbicidal effect of UV light depends on the dose in both, disinfection processes and natural inactivation by the sunlight in surface water. Deviations of the time dose reciprocity are well known from chemical water disinfection whereas no data are available about this effect in UV inactivation in water. In a previous study we found that the UV inactivation behaviour of yeast strains does not follow the time dose reciprocity, insofar that longer exposure led to higher reduction of cultivable cells. In contrast, an earlier study about E coli B/r claimed a higher inactivation with single exposure compared with fractionated UV irradiation. To investigate this question we selected water-relevant microorganisms and studied their UV inactivation behaviour (253.7nm) by means of a specially designed UV irradiation apparatus (a) under standard irradiation conditions (2W/m2) and (b) with three levels of UV dose rate (2, 0.2 and 0.02W/m2). The test organisms were (i) three E coli strains (ATCC 25922, ATCC 11229 and an isolate from sewage) representing the routinely used faecal indicator, (ii) three bacterial viruses (MS2, ϕX174 and B40-8) proposed as indicators for viral contamination in water and (iii) spores of Bacillus subtilis because of their use as a biodosimeter in prototype testing of commercial UV plants for drinking water disinfection. We found, under standard inactivation conditions, that the E coli strains and phage ϕX174 are most UV susceptible, followed by B40-8 and finally MS2 and bacterial spores. The dose protraction experiments revealed for the E coli strains a higher inactivation with high dose rates compared to low dose rates at the same UV doses (difference of about 1 log10 at 80-100J/m2). The other test organisms did not deviate from the time dose reciprocity in the proven range of dose.