Objective. To analyze clinical and fMRI indicators related with the therapeutic effectiveness of complex restorative treatment including intensive speech therapy in various clinical forms and severity of the aphasia syndrome. Material and methods. The study included 40 right-handed patients with aphasia syndrome after the first hemispheric ischemic stroke. Patients were studied 3 month after disease onset. Patients were examined before and after rehabilitation treatment (4.7 weeks) including intensive speech therapy (15 hours of exercise per week). Evaluation of the effectiveness of treatment was carried out based on the dynamics of the results of neuropsychological, neurological and neuroimaging examinations of patients before and after treatment. The degree of recovery of speech and non-speech cognitive functions (based on the 10-point cognitive assessment, the dynamics of focal neurological deficit (NIHSS) and functional recovery of patients (Barthel index, modified Rankin scale) were assessed. Neuroimaging methods included structural MRI and two fMRI methods: fMRI resting state and fMRI equivalent of the evoked potentials of mismatch negativity (HP), using a sequence of standard and deviant sounds (Russian phonemes) to obtain the equivalent of HP. For statistical calculations, the program in Python version 3.6.0 was used. Results. The maximum efficacy of restorative treatment (≥15% improvement in the scores of 10-point quantifying) was detected in 28 (70%) cases: 18 patients with initially severe and 10 patients with moderate aphasia, regardless of the clinical form of aphasia, and 11 patients with isolated sensory aphasia. Regression of speech and non-speech cognitive impairments in sensory aphasia was accompanied by a pronounced activation of intact speech homologues of the right hemisphere and the appearance of a small temporal and parietal region, which was small in volume and weak in intensity, similar in the localization to the norm. Regardless of the clinical form of aphasia, maximum therapeutic efficacy was accompanied by a reorganization of the speech neuronal rest network, which includes enhancement of both intra- and interhemispheric functional connectivity, with the predominance of amplification of the intra-hemispheric interactions of the posterior speech zones to the left while reducing their interhemispheric connectivity. Significant improvement in motor and sensory functions was detected in 9 (22%) patients with moderate contralateral spastic hemiparesis (p≤0.00) and did not correlate with the degree of regression of speech and non-speech focal cognitive impairments. A slight therapeutic effect was observed in 12 (30%) patients with mild to moderate speech and non-speech cognitive impairment. This effect was not correlated with a specific clinical form of aphasia. Conclusion. The maximum effectiveness of the 4.7 week restorative inpatient treatment, including intensive speech therapy (15 sessions per week), was obtained mainly with severe aphasia, and also with a specific clinical form — isolated sensory aphasia. According to the fMRI data, various compensatory reorganization of the neural speech network was detected, which probably reflects post stroke neuroplasticity associated with both the severity of aphasia and its specific clinical form.