Abstract The impact of UV-B radiation (UVBR; 280–320 nm) on lower levels of a natural plankton assemblage (bacteria, phytoplankton and microzooplankton) from the St. Lawrence Estuary was studied during 9 days using several immersed outdoor mesocosms. Two exposure treatments were used in triplicate mesocosms: natural UVBR (N treatment, considered as the control treatment) and lamp-enhanced UVBR (H treatment, simulating 60% depletion of the ozone layer). A phytoplankton bloom developed after day 3, but no significant differences were found between treatments during the entire experiment for phytoplankton biomass (chlorophyll a and cell carbon) nor for phytoplankton cell abundances from flow cytometry and optical microscopy of three phytoplankton size classes (picoplankton, nanoplankton and microplankton). In contrast, bacterial abundances showed significantly higher values in the H treatment, attributed to a decrease in predation pressure due to a dramatic reduction in ciliate biomass (∼70–80%) in the H treatment relative to the N treatment. The most abundant ciliate species were Strombidinium sp., Prorodon ovum and Tintinnopsis sp.; all showed significantly lower abundances under the H treatment. P. ovum was the less-affected species (50% reduction in the H treatment compared with that of the N control), contrasting with ∼90% for the other ones. Total specific phytoplanktonic and bacterial production were not affected by enhanced UVBR. However, both the ratio of primary to bacterial biomass and production decreased markedly under the H treatment. In contrast, the ratio of phytoplankton to bacterial plus ciliate carbon biomass showed an opposite trend than the previous results, with higher values in the H treatment at the end of the experiment. These results are explained by the changes in the ciliate biomass and suggest that UVBR can alter the structure of the lower levels of the planktonic community by selectively affecting key species. On the other hand, linearity between particulate organic carbon (POC) and estimated planktonic carbon was lost during the postbloom period in both treatments. On the basis of previous studies, our results can be attributed to the aggregation of carbon released by cells to the water column in the form of transparent exopolymer particles (TEPs) under nutrient limiting conditions. Unexpectedly, POC during such a period was higher in the H treatment than in controls. We hypothesize a decrease in the ingestion of TEPs by ciliates, in coincidence with increased DOC release by phytoplankton cells under enhanced UVBR. The consequences of such results for the carbon cycle in the ocean are discussed.