The study aims to assess the feasibility of using a computer assisted diagnostic interview by GPs and to examine the level of agreement between the Spanish version GMHAT/PC diagnosis and psychiatrists' ICD-10 based clinical diagnosis. Participants in the study ranged from those who were in remission to others who had different mental illnesses. They were recruited from inpatient and outpatient mental health settings. All consecutive patients were interviewed using Spanish version of GMHAT/PC and they were assessed independently by psychiatrists to in order to get their ICD-10 based diagnosis. Two hundred ninety-nine patients participated in the study. The mean duration of interview was 12.5 minutes. There is an acceptable to good level of agreement between the GP’s (GMHAT/PC) diagnoses and the psychiatrists’ (clinical) diagnoses of any mental illness, Kappa 0.58 95% C.I (0.46, 0.72). There is good level of sensitivity (81%) and specificity (92%), with GPs correctly identifying 242 out of the 250 participants diagnosed with mental illness and 27 out of 35 of those without. The finding of the study suggest that GMHAT/PC Spanish version used by GPs detected mental disorders accurately and it was feasible to use GMHAT/PC (Spanish version) in Latin America settings.