Prior to the spring bloom in 2003 and 2004, batch temperature experiments of approximately 3 weeks' duration were carried out in land-based mesocosms in at the Espeland field station (Norway), with temperatures on average increased ~ 2.7-3 °C (T1) and ~ 5.2-5.6 °C (T2) above in situ fjord temperature (RM). The development in the chlorophyll concentrations showed an earlier bloom as a response to increased temperatures but the carbon biomass showed that the warmest treatment yielded the lowest biomass. This study indicates that a part of the relationship between temperature and spring bloom timing stems from a temperature-induced change in phytoplankton algal physiology (the efficiency of photosystem II, F_v/F_m, and growth rates, µ_max), i.e. a direct temperature effect. Data analysis performed on microscope identified and quantified species did not show a significant temperature influence on phytoplankton community composition. However, the HPLC data indicated that temperature changes of as little as 3 °C influence the community composition. In particular, these data showed that peridinin-containing dinoflagellates only increased in abundance in the heated mesocosms and that a prasinophycean bloom, which was undetected in the microscope analyses, occurred prior to the blooms of all other phytoplankton classes in all treatments. The microscope analyses did reveal a temperature effect on individual species distribution patterns. Thalassionema nitzschioides was more abundant in the warm treatments and, in the warmest treatment, the spring bloom forming Skeletonema marinoi comprised a smaller proportion of the diatom community than in the other treatments.