The relative efficacy of Neuro-Developmental Treatment (N.D.T.) versus the Brunnstrom method was studied from the perspective of the functional recovery of stroke patients. An alternating treatment design (B-C-B-C) was used. Each intervention phase lasted 5 weeks. The functional recovery of the patients was assessed every week by using the Barthel Index and the Action Research Arm test, by registering walking velocity, and by performing gait analysis. At the start and at the end of each intervention phase, neurological and neuropsychological assessments took place. Time series analysis indicated that for one patient only, walking speed progressed more during the Brunnstrom phases than the N.D.T. phases. This result did not generalize towards other parameters or patients studied. For upper extremity function and maximum walking speed, a high correlation was found between the week post stroke in which the patient showed the first signs of recovery and the end condition after 20 weeks.