Software Effort Estimation is a software estimation process as an essential process in software projects. Many studies have been carried out regarding software estimation with various methods, including machine learning methods or non-machine learning methods. This research was conducted using a public data set and carried out through a Particle Swarm optimization attribute selection experiment on the project parameters using the K-NN algorithm as an estimate. The software estimation effort dataset used in this study is Desharnais, Maxwell, Kitchenham CSC, and Kamrer. The results of the study indicate that the Particle Swarm optimization feature selection can reduce the RMSE value and show a significant decrease in the RMSE value. This shows that the lower the RMSE value, the more precise the estimated value will be. This research is useful where it can be used as information for developers in estimating the time to build an application so that it is in accordance with the initial plan.