A workshop with the aim to compare classical and molecular techniques for phytoplankton enumeration took place at Kristineberg Marine Research Station, Sweden, in August 2005. Seventeen different techniques – nine classical microscopic-based and eight molecular methods – were compared. Alexandrium fundyense was the target organism in four experiments. Experiment 1 was designed to determine the range of cell densities over which the methods were applicable. Experiment 2 tested the species specificity of the methods by adding Alexandrium ostenfeldii, to samples containing A. fundyense. Experiments 3 and 4 tested the ability of the methods to detect the target organism within a natural phytoplankton community. Most of the methods could detect cells at the lowest concentration tested, 100 cells L−1, but the variance was high for methods using small volumes, such as counting chambers and slides. In general, the precision and reproducibility of the investigated methods increased with increased target cell concentration. Particularly molecular methods were exceptions in that their relative standard deviation did not vary with target cell concentration. Only two of the microscopic methods and three of the molecular methods had a significant linear relationship between their cell count estimates and the A. fundyense concentration in experiment 2, where the objective was to discriminate that species from a morphologically similar and genetically closely related species. None of the investigated methods were affected by the addition of a natural plankton community background matrix in experiment 3. The results of this study are discussed in the context of previous intercomparisons and the difficulties in defining the absolute, true target cell concentration.