The mapping of cognitive trajectories after a first episode of schizophrenia has been the aim in several studies, but the longitudinal course of cognitive impairments remains an important question. Due to methodological limitations, it has been challenging to pinpoint specific periods of improvement or stability in cognitive functioning over time. The objective of this study is to further clarify the longitudinal course of cognitive change after a first episode of schizophrenia through frequent repeated measurement. A total of 56 persons participated in the study (28 first episode patients and 28 healthy pairwise matched controls) with 79 % of patients retained at the 10-year follow-up. The Oslo Schizophrenia Recovery study has a repeated measurement design and includes data from nine cognitive assessments over 10 years. Cognition was assessed with the MATRICS Consensus Cognitive Battery, which is well suited for repeated measurements. Data were analyzed with linear multilevel models. The results challenge some of the views about the course of cognitive impairment in first-episode schizophrenia patients. Using quadratic time effects in our analyses and balancing the patient group with regards to the most relevant confounding demographic variables such as age, gender, and education, we showed that cognitive deficits change and improve more than in healthy individuals until year 6, when both groups stabilize. The patient group improved on some of the most important cognitive domains associated with functional outcome with 63.5 % full recovery at 10-year follow-up.