Objectives: How brain damage after stroke is related to specific clinical manifestation and recovery is incompletely understood. We studied cognitive reserve (CR) in stroke patients by two types of measurements: (i) objectively verifiable static proxies (i.e., education, occupational attainment), and (ii) subjective, dynamic proxies based on patient testimony in response to a questionnaire. We hypothesized that one or both of these types of CR measurements might correlate positively with patient cognitive performance during the post-acute and chronic phases of recovery. Method: Thirty-four stroke patients underwent neuropsychological assessment at 2, 6 and 24 months after stroke onset. In chronic stage at 24+ months, self-rating assessments of cognitive performance in daily life and social integration were obtained. CR before and after stroke was estimated using static proxies and dynamic proxies were obtained using the Cognitive Reserve Scale (CRS-Pre-stroke, CRS-Post-stroke). Results: CRS-Pre-stroke and CRS-Post-stroke showed significant mean differences. Dynamic proxies showed positive correlation with self-assessment of attention, metacognition, and functional ability in chronic stage. In contrast, significant correlations between static proxies and cognitive recovery were not found. Conclusions: Dynamic proxies of CR were positively correlated with patients’ perception of their functional abilities in daily life. To best guide cognitive prognosis and treatment, we propose that dynamic proxies of CR should be included in neuropsychological assessments of patients with brain damage.