The current standard for testing tetrodotoxin (TTX) in foodstuffs is the mouse bioassay (MBA) in Korea as in many other countries. However, this test suffers from potential ethical concerns over the use of live animals. In addition, the mouse bioassay does not test for a specific toxin thus a sample resulting in mouse incapacitation would need further confirmatory testing to determine the exact source toxin (e.g., TTX, STX, brevotoxin, etc.). Furthermore, though the time of death is proportional to toxicity in this assay, the dynamic range for this proportional relationship is small thus many samples must be diluted and new mice be injected to yield a result that falls within the quantitative dynamic range. Therefore, in recent years, there have been many efforts in this field to develop alternative assays. High performance liquid chromatography (HPLC) coupled with mass spectrometry (MS) has been emerged as one of the most promising options. A LC-MS-MS method involves solid-phase extraction (SPE) and followed by analysis using an electrospray in the positive ionization mode and multiple reactions monitoring (MRM). To adopt LC-MS-MS method as alternative standard for testing TTX, we performed a validation study for the quantification of TTX in puffer fish. This LC-MS-MS method showed good sensitivity as limits of detection (LOD) of 0.03~0.08 μg/g and limits of quantification (LOQ) of 0.10~0.25 μg/g. The linearity (r2) of tetrodotoxin were 0.9986~0.9997, the recovery were 80.9~103.0% and the relative standard deviations (RSD) were 4.3~13.0%. The correlation coefficient between the mouse bioassay and LC/MS/MS method was higher than 0.95.