This paper describes human identification using fisherface method to identify someone. The output is whether recognized or not an input image as an individual in the database. There are four main stages for this method, mainly face detection, PCA (Principal Component Analysis) calculation, FLD (Fisher's Linear Analysis) calculation and classification stage. In face detection stage, color thresholding is used to segment pixels that contain skin color. PCA calculation and FLD calculation stages are used to form a set of fisherfaces from a training set or database that will be used. All face images can be reconstructed from the combination of fisherfaces with different weights for each face image. The last stage, classification stage, is to identify the input image by comparing the weight of fisherface required to reconstruct the input face towards face images in the training set. The weight calculation is done by using Euclidian distance method. The simulations are done for 66 input images and the successful recognition rate is about 81.82%.