Website counterfeit or phishing website is a crime in the virtual world whose popularity is increasing even in Indonesia until now in 2015. In this study we take on phishing websites dataset from UCI Repository as many as 2546 data by 30 variables used to determine the website is a phising website or not. Having obtained the data, the authors conducted a study to determine the most appropriate algorithms. Determination of the algorithms with comparisons between algorithms classification techniques. Based on some related research and the advantages of the algorithm, the authors took five algorithms to be tested, the algorithm Decission Tree (C4.5), Naive Bayes, KNN, Support Vector Machine and Neural Network. This study using a test of accuracy and AUC as well as different test parametric T-test. In each model, the authors divide the main data into five sections, and on each of the training data validation was done using K-Fold Cross Validation. The results of this study demonstrate that Neural Network algorithm and SVM into the most appropriate algorithm used by the average value of accuracy is 94 and the value AUC 0.9.