Red tides by the ichthyotoxic dinoflagellate Cochlodinium polykrikoides have caused large scaled mortality of fish and great loss in aquaculture industry in many countries. Detecting and quantifying the abundance of this species are the most critical step in minimizing the loss. The conventional quantitative real-time PCR (qPCR) method has been used for quantifying the abundance of this species. However, when analyzing >500 samples collected during huge C. polykrikoides red tides in South Sea of Korea in 2014, this conventional method and the previously developed specific primer and probe set for C. polykrikoides did not give reasonable abundances when compared with cell counting data. Thus improved qPCR methods and a new specific primer and probe set reflecting recent discovery of 2 new ribotypes have to be developed. A new species-specific primer and probe set for detecting all 3 ribotypes of C. polykrikoides was developed and provided in this study. Furthermore, because the standard curve between cell abundance and threshold cycle value (Ct) is critical, the efficiencies of 4 different preparation methods used to determine standard curves were comparatively evaluated. The standard curves were determined by using the following 4 different preparations: (1) extraction of DNA from a dense culture of C. polykrikoides followed by serial dilution of the extracted DNA (CDD method), (2) extraction of DNA from each of the serially diluted cultures with different concentrations of C. polykrikoides cultures (CCD method), (3) extraction of DNA from a dense field sample of C. polykrikoides collected from natural seawater and then dilution of the extracted DNA in serial (FDD method), and (4) extraction of DNA from each of the serially diluted field samples having different concentrations of C. polykrikoides (FCD method). These 4 methods yielded different results. The abundances of C. polykrikoides in the samples collected from the coastal waters of South Sea, Korea, in 2014–2015, obtained using the standard curves determined by the CCD and the FCD methods, were the most similar (0.93–1.03 times) and the second closest (1.16–1.33 times) to the actual cell abundances obtained by enumeration of cells. Thus, our results suggest that the CCD method is a more effective tool to quantify the abundance of C. polykrikoides than the conventional method, CDD, and the FDD and FCD methods.