Deception involves emotions of fear and guilt. These negative emotions are expressed in language in terms ofpsychological distance from the deception object. The psychological distance and emotional experience reflect an attemptto control the negative mental representation. More especifically emotional distance is represented in deceptive language bymeans of cues of reference, verb tense and detail avoidance. Then, hints of emotions of fear and guilt should be displayedin language.The present work analyses emotional language cues for deception detection by means of Machine Learning(ML)techniques and Linguistic Inquiry and Word Count (LIWC). Results show that Support Vector Machines (SVM) best representsthe discrimination between true and false information (up to 74.15 % of accuracy rates) based only on the effect of emotion indeceptive speech.