Large subunit ribosomal DNA (LSU rDNA) sequences have been increasingly used to infer the phylogeny and species identity of organisms, a few previous studies, however, have observed high intraspecific and even intraindividual variability in LSU rDNA in some dinoflagellate species due to, assumably, large copy numbers of rDNA in dinoflagellates. Since the copy number of LSU rDNA varies tremendously among dinoflagellate species, the intraspecific and intraindividual diversity for a species of particular interest thus needs to be investigated individually. As a toxic and HABs-forming dinoflagellate, Margalefidinium (= Cochlodinium) fulvescens has been observed to approach blooming density in Jiaozhou Bay, China since 2015 after numerous blooms having been reported from other countries. In trying to identify the source of this newly observed HABs-forming species in China by sequencing the LSU rDNA for both field samples and clonal cultures, we noticed and thus further investigated high intrapopulational and intraindividual genetic diversities of the dinoflagellate. The D1–D6 region of the LSU rDNA (1,435 bases) was amplified from 7 field samples (pooled cells) and 11 clonal cultures, cloned, sequenced, and analyzed phylogenetically for 2,341 sequences obtained. All the numbers of sequences obtained from each clonal culture were far less than the estimated rDNA copy number in M. fulvescens. In the clone library, only one unique sequence was contained in all samples as the most dominant sequence. We found high intrapopulational and intraindividual genetic diversity in M. fulvescens as reflected in the number of polymorphic sites and unique sequences in the clone library for different field samples and clonal cultures in comparison to other species. The mean number of nucleotide differences of each sequence from different field samples and clonal cultures were 6.43 and 4.42 bases, respectively, with the highest being 132 bases, nearly 10%. The sequences with highest variability may be easily annotated as different species if they were obtained from environmental genomic studies because sequence-based species identification in meta-barcoding studies often use "97% identity" threshold. Based on that the mean and overall intrapopulational genetic diversity calculated for 7 field samples was equivalent to the mean and overall intraindividual variability for 11 clonal cultures in indices of genetic diversity, together with the result of AMOVA analysis, we infer that the variability within individual cells (i.e. variability among LSU rDNA polymorphic copies) caused both the intraindividual and intrapopulational genetic diversities observed in the M. fulvescens population, and a higher interpopulational diversity may exist among different geographic populations. The results provide an insightful basis for such a comprehensive interpopulational comparison and important implications for identifying species and establishing new taxa based on the similarity comparison to reference sequences deposited in databases.