This study deals with a Turkish text analysis system for adults to detect psychiatric disorders. The data is organized as answers of two questions for each subject. The aim of this analysis system is to determine whether the subjects are healthy or suffering from a psychiatric disorder, namely Depression or Anxiety based on their language usage. Naïve Bayes, Support Vector Machine and Decision Tree ML methods are used. Words or categorized words are used as features. The aim of this research is presenting a psychiatric disorder automatic classification system using the ML methods as one of the initiators of Turkish natural language processing researches.