Tomatoes have a risk of carrying pesticides above the maximum residue limit (MRL) because the fruit is directly sprayed with pesticides during its production process. Pesticide residue in farmers’ produce pose indirect effects to the consumers, but in the long run, it may cause health problems such as neural disorders as well as enzyme metabolism. This research identifies the image of tomatoes containing pesticides by using two types of tomatoes were used as samples, namely tomatoes which contain pesticides, and those which do not contain pesticides. This research aims to develop an algorithm to identify tomatoes that contain pesticides and those which do not contain pesticides using Learning Vector Quantization (LVQ). The characteristics used to identify tomato images are average, variant, and standard deviation. This research consisted of two classes and used 40 training image data and 40 test image data for each class. During the training process using LVQ parameters, there were 98.75% best percentage at alpha 0.001 and decalpha 0.9 with the lowest iteration of 3. The final weight obtained from the parameters was then used to perform test data identification. In terms of the best performance on the test data, it was with alpha 0.001 and decalpha 0.9, which reached 97.5%.