This study was conducted to classify the quality of employee performance, the results of which can be used by the Central Bureau of Statistics in Samarinda City to determine the clusters of each employee. The quality evaluation of the employee's performance is carried out annually. The assessment results are grouped into several clusters as consideration for leaders to provide job recommendations to their employees. The clustering was applied using the k-means algorithm. Data clustering was performed based on the closest distance to the center of the cluster. In this study, three algorithms were applied to determine the distance to the center of the centroid to see the comparison. The sample data used were 25 employees with five attributes: professionalism, integrity, trustworthiness, employee performance achievements, and absenteeism. The data were clustered into 3 clusters, which are the most optimal number of clusters based on the Sum of Square test results with a value of 3.55, which is the value with the enormous difference. The results of the implementation of the clustering method were obtained: 12 employees are in cluster one, 10 employees are in the second cluster, and 3 employees are in the third cluster. Based on the value of the centroid center in the last iteration, it was concluded that the employees in the first cluster were the employees with the best scores, the second cluster were the employees with the medium scores, and the third cluster was the employees with the lowest scores.