In recent decades, the frequency, intensity and geographical distribution of Harmful Algal Blooms (HAB) have increased. Harmful Algal Blooms (HAB) not only harms Marine fishery and aquaculture industry, deteriorates Marine environment, affects coastal tourism, but also brings many problems to human health. Therefore, we need to pay more attention to the coastal detection in order to reduce or prevent such impacts. At present, in addition to the basic microscopic identification methods, molecular methods are more powerful tools. However, as ribosomal RNA(rRNA) genes are the target of molecular identification, the number and proportion of each species of algae are determined according to the number of rRNA genes. However, it is known that the copies number of rRNA genes in different algae is different. This requires us to accurately quantify the copies numbers of rRNA genes of different algae, and then normalize them into molecular data to obtain a more realistic number and proportion of algae. In this study, the copy number of rRNA genes in single cells of 25 species of algae was quantified by digital PCR(dPCR). Chelex buffer found to be suitable for DNA extraction, which could obtain DNA in one step and avoid the loss caused by transfer to underestimate the copies number. Finally, the copies number data were compared with high-throughput data, and the number and proportion of known algae were used for verification. The absolute quantification of 18S rRNA gene copies number per cell in 25 phytoplankton species was successfully demonstrated by dPCR.