Speech is an active source to prompt emotions and attitude through language. Discovering the emotional content within a speech signal and then recognizing the type of emotion developed from the uttered speech is a significant task for researchers. In this paper, we present a study of emotion recognition in Arabic speech using the KSUEmotions Arabic speech emotion corpus by applying feature-extraction techniques followed by classification techniques like the Knearest neighbor algorithm (KNN) and support vector machine (SVM). The experiments were performed using the Python programming language. The experiments revealed that KNN is better than SVM for this corpus, and the results of the experiment show the highest accuracy for the emotion of sadness, followed by happiness and then by surprise.