Automatically analyzing the user's emotion from his/her texts has been gaining interest as a research field. Emotion classification of English texts is studied by several researchers and promising results have been achieved. In this work, an emotion classification study on Turkish texts is presented. To the best of our knowledge, this is the first study conducted on emotion classification for Turkish texts. Due to the nature of Turkish language, several pruning tasks are applied and new features are constructed in order to improve the emotion classification accuracy. We compared the performance of several classification algorithms for emotion analysis and reported the results.