Objective: To evaluate the varying effects of rehabilitation intensity classified by the number of treatment sessions on recovery of activity and function in stroke subjects. Method: Eighty nine subjects with stroke (51 infarction, 38 hemorrhage) had received conventional rehabilitation programs (physical therapies, occupational therapies) with 30 minutes per each therapy. They were divided into two groups; group I consisted of 42 subjects who received one session of rehabilitation therapy per day; group II consisted of 47 subjects received two sessions per day. Functional outcomes, such as K-BBS (Korean version of Berg Balance Scale), FIM (Functional Independence Measure), and MMSE-K (Korean version of Mini Mental State Examination) were assessed with two weeks interval. Data was analyzed the differences of functional outcomes assessed at the initial time of treatments (the initial) and at the time of the peak K-BBS (the second). Results: Age, sex, lesion sites, the initial K-BBS, the initial FIM, and MMSE-K scores had no differences between two groups (p>0.05). There were significant differences between the initial and the second K-BBS, and the initial FIM and the second FIM in group I & II, respectively (p 0.05). Also, length of stay was no significant differences between two groups. Conclusion: The above findings suggest that though all rehabilitation programs affect the functional improvement of stroke subjects, the daily rehabilitation training time could be more important factor for functional improvement in subjects with stroke, especially on recovery of ambulation rather than the sum of functional gain. (Brain & NeuroRehabilitation 2009; 2: 134-139)