The model built using the data mining process can be used to make predictions from the data. The model can be built using a datasheet that contains data that is processed from the process. One implementation of the model in data mining is the prediction of a disease such as diabetes. In this study, a data mining model was developed using the k-NN algorithm and data normalization was carried out. The normalization method used is Z-Score and Min-Max. The research methodology is carried out by first determining the datasheet, selecting the data mining model and dividing the datasheet into datasheets into training data and data testing and evaluating the performance of the model created. The process of modeling using python programming. The data mining process uses a classification model using the k-NN algorithm. The datasheet used is a public datasheet, namely the diabetes datasheet which consists of 768 records and 8 attributes. The results of this modeling show that the normalization process can provide better accuracy values. The model developed without normalization produces a value of k=5 with an accuracy of 70%, normalization with the Z-Score method produces a value of k=21 with an accuracy of 72%, normalization with Min Max produces a value of k=3 with an accuracy of 74%. The recommended model is k-NN mode with a value of k=3.