Phytoplankton blooms, in their pivotal position in pelagic seasonal succession require precise classification criteria in order to evaluate such parameters as bloom start, bloom timing, bloom maximum and growth rates. Such bloom parameters are linked directly to species and bloom specific features. Currently the phytoplankton bloom concept, though intuitively clear, lacks operational criteria allowing the precise definition of bloom parameters. We present a semi-quantitative method of classification of marine phytoplankton blooms based on an algorithmic estimation of several bloom descriptors computed from densely recorded phytoplankton data, like the Helgoland Roads long-term time series. Combining these descriptors we propose a novel classification scheme which may serve useful in the discussion of species fitness, competition and succession of marine algae. Special emphasis is put on the detection of the bloom start, because of its crucial importance for many current research topics, including trigger mechanisms and climate-induced temporal shifts in the context of the match/mismatch hypothesis. Visual examination of scatter plots of these parameters leads us to propose three types of blooming algae. © 2009 Elsevier B.V. All rights reserved.