Abstract: ObjectiveTo analyze the spatial distribution pattern of sampling locations of food safety risk surveillance.MethodsData were collected from 281 sampling sites of vegetables and their products for food safety risk surveillance of 22 monitoring cities in a province in 2015. The sampling location data were converted into latitude and longitude coordinates by GIS Geocoding technology, and then, a sampling location thematic map was created. After calculating the average nearest neighbor distance of the sampling location, the nearest neighbor index (NNI), namely the ratio of the average nearest neighbor distance to the expected value of the nearest neighbor distance of sampling location in each monitoring city, was calculated to analyze the pattern of the spatial distribution pattern.ResultsThe NNIs of the sampling locations of the 22 monitoring cities ranged from 0.002-1.086, with an average NNI of 0.335. The NNIs were less than 0.1 in 18% (4/22) monitoring cities, ranged from 0.1 to 0.5 in 68%（15/22）cities, ranged from 0.5 to 1 in 9%（2/22）cities, and were bigger than 1 in 5% (1/22) cities. Spatial randomness test showed that only one city's P value was greater than 0.05 (NNI=1.086，Z=0.57，P=0.28), while the NNIs of the remaining 21 cities were less than 1, P<0.05.ConclusionThe samples of these cities were neither random nor independent, and the bias would be caused using the classical statistical method to infer the overall pollution situation. The nearest neighbor spatial analysis method can analyze the spatial distribution pattern of sampling locations in different cities, and find the cluster sampling locations, to avoid biased sample data and improve the scientific level of sampling scheme for food safety risk monitoring.