Introduction: The purpose of the present research is aimed at studying the relationship between emotional intelligence as an ability and emotional intelligence as a trait and mental health of a sample of school counsellors. Method: The sample has been made up of 203 school counsellors. The instruments used have been: Mayer-Salovey-Caruso Emotional Intelligence Test (MSCEIT), Trait Emotional Intelligence Questionnaire Short Form (TEIQUE-SF) and the Mental Health Scale (MH-5). In the first place, the correlations among the scoring corresponding to all of the variables were analysed. Subsequently, the differences in average (t of Student) were studied according to the educational stage in which the school counsellors are working and according to their sex. Finally, with the aim of exploring the variables that predict mental health, an analysis of multiple lineal regression was carried out, step by step. Results: The results show that the existing relationships between EI as a trait and mental health are higher than those existing between EI as an ability and mental health. On the other hand, scorings significantly higher were registered in the school counsellors in Secondary School in the variable of emotional intelligence as an ability in emotional comprehension. Scorings significantly higher are also registered in the variable of mental health of school counsellors in Primary Education. Likewise, scorings significantly higher are also registered in women in their score of emotional intelligence as an ability in emotional regulation, and scorings significantly higher in the score of emotional intelligence as a personality trait in men. Finally, the analysis of multiple regression has allowed for identifying emotional intelligence as a trait as a predictor variable of mental health. Discussion: The present study represents one of the first research papers carried out in a sample of school counsellors in which the influence on mental health of emotional intelligence as a trait and as an ability is analysed. Furthermore, it provides evidence about the predictive validity of EI as a trait over an indicator of mental health.