OBJECTIVE To establish a method for simultaneous determination of the contents of 6 kinds of N-nitrosamines genotoxic impurities in losartan potassium raw material and its formulations. METHODS GC-MS/MS was adopted to determine 6 kinds of N-nitrosamines genotoxic impurities in losartan potassium raw material， Losartan potassium tablet， Losartan potassium capsule and Losartan potassium hydrochlorothiazide tablets， such as N-nitrosodimethylamine （NDMA）， N-nitrosodiethylamine （NDEA）， N-ethyl-N-nitroso-2-propanamine （NEiPA）， N-nitrosodiisopropylamine （NDiPA）， N-nitrosodipropylamine （NDPA） and N-nitrosodibutylamine （NDBA）. The separation was performed on SHIMADZU SH-L-17Sil MS capillary column by temperature- programmed GC， with injector temperature of 250 ℃ ， sample size of 1 μL， carrier gas of helium， and carrier flow rate of 1 mL/min. Electron ionization and multiple reaction monitoring （MRM） data acquisition mode were used， with an ion source temperature of 250 ℃ and solvent delay time of 3.1 min. RESULTS The separation among NDMA， NDEA， NEiPA， NDiPA， NDPA， NDBA and adjacent chromatographic peaks was good， and the separation rate was higher than 3.8； the linear ranges of them were 4.9-486.0， 4.9-488.5， 4.5-451.5， 6.8-683.5， 5.2-525.0 and 5.2-520.0 ng/mL（all r≥0.999 8）. The limits of quantitation were 4.86， 4.88， 4.52， 6.84， 5.25 and 5.20 ng/mL； the limits of detection were 0.97， 0.98， 0.90， 1.37， 1.05 and 1.04 ng/mL. RSDs of repeatability tests were 2.2%-5.6%（n＝6）， those of precision tests were 0.5%-1.4%（n＝6）， and those of stability tests were 1.5%-3.4%（n＝5）， respectively. Average recoveries of low-， medium- and high-concentration solution were 83.4%-103.0% （RSDs were 1.2%-6.3%， n＝3）， respectively. No one among the 6 kinds of N-nitrosamines genotoxic impurities was detected in both losartan potassium raw material and formulations. CONCLUSIONS The method is good in separation effect， highly accurate， sensitive and simple. It can be used in the determination of the 6 kinds of N-nitrosamines genotoxic impurities.