Background: The Rorschach Inkblot Method is regarded as an important clinical instrument for detailed diagnostic description of the integrative capacities of individuals in psychotic states and as an instrument for measuring progression in the course of treatment. Aims: To describe relevant Rorschach variables at baseline in a group of consecutively admitted patients with first-episode schizophrenia. Furthermore, to describe the changes in these variables from baseline to year 2 for the group of patients given psychiatric standard treatment, and to compare these changes with changes in other outcome measures [Positive and Negative Syndrome Scale (PANSS), Global Assessment of Functioning (GAF), Strauss-Carpenter and socio-demographic variables]. Methods: In a prospective study, 34 patients consecutively admitted to treatment for a first episode of schizophrenia were tested using Exner's Rorschach Comprehensive System at inclusion and after 2 years. Results: Core variables of the Rorschach, assumed to show changes (e.g. reality testing, perceptual and thought disorders) in patients with a first episode of schizophrenia, all improved but revealed no significant changes after 2 years of treatment, while other measurements (PANSS, GAF) did. Furthermore, discrepancies were found between the Perceptual-Thinking Index (PTI) and PANSS(positive) symptoms of delusions and hallucinations. Conclusion: The small sample limits the conclusions, but the Rorschach variables may have problems reflecting major changes in psychotic symptoms and social functioning. With short protocols, PTI seems at risk of being unnecessarily insensitive and PTI may benefit from changes in the scoring procedures.