Microcystin-producing cyanobacteria blooms, as well as their increasing global occurrence, are of worldwide concern because they have been implicated in the death of animals and humans. Currently, many techniques are used to evaluate whether toxins are present in water samples, including high-performance liquid chromatography (HPLC), mouse bioassays, protein phosphatase inhibition assays (PPIA), enzyme linked immunosorbent assay (ELISA), and mass spectrometry. While some of these techniques are quite sensitive, they are of use only once the toxin is present in the water source. However, microcystins remain within cyanobacterial cells until being released in high concentrations during cell lysis. Hence, identifying potential microcystin-producing strains in the environment prior to the release of toxins into the water body is more useful than detection of toxins. Many microcystin-producing strains cannot be differentiated from non-toxic strains by traditional microscopy or with ribosomal gene sequences. Synthetase genes (mcy) are responsible for the biosynthesis of microcystin and are only present in the genome of toxic strains. Polymerase chain reaction (PCR) assays based on mcy gene sequences have been reported. Though these assays for rapid, sensitive, and specific detection appear to be promising, PCR applied as a rapid diagnostic test in the field has been seriously constrained owing to the complexity of the amplification conditions and the need for specialized high-cost instruments. Recently, loop-mediated isothermal amplification (LAMP), a new technology for amplifying DNA, has been developed. Unlike PCR, which requires special instruments to control reaction temperature, LAMP requires only a temperature-controlled water bath. In this manuscript, we designed a set of LAMP primers to target the sequence of the mcyG gene, which is responsible for the first step in the biosynthesis of Adda (3-amino-9-methoxy-2, 6, 8-trim ethyl-10-phenyldeca-4, 6-dienoic acid), an important part of microcystin. The conditions of amplification were optimized to implement the sensitivity and specificity experiments. We found that 61 ℃ for 40 min was the most suitable reaction condition. We then selected thirteen cyanobacterial strains, including Microcystis, Aphanizomenon, Chroococcus, Nostoc, Oscillatoria, and Anabaena, to determine the specificity and sensitivity of the primers in the LAMP reactions. The results of amplification were positive with ten microcystin-producing strains and negative with three non-toxic strains, which demonstrated that the primers had suitable levels of specificity. The sensitivity of LAMP, with a detection limit of 24 cfu/ml, was 1000-fold higher than that of PCR. For verifying the application of this LAMP assay in the aquatic ecosystem, seven environmental samples from ponds and lakes in Ningbo and one from Tai Lake were analyzed using both the LAMP assay targeting the mcyG gene and a PCR assay. The LAMP assay detected amplification in two water samples that the PCR assay did not detect, demonstrating that the LAMP assay is more effective and more suitable for field testing. The LAMP method is fast, simple, and low in cost making it advantageous for the detection of potential toxic cyanobacterial blooms. By employing the LAMP assay, workers at the Environmental Monitoring Station will be able to detect toxic cyanobacterial cells at low levels, before they reach an unsafe concentration. This will allow the workers, through routine monitoring of potential toxin status in the freshwater, to have enough time to implement a series of control strategies for avoiding human exposure to cyanobacterial hepatotoxins at dangerous levels.