The aim of this study was to evaluate the effects of walking on a treadmill while using dynamic functional electrical stimulation (Dynamic FES) on functional ability and gait in chronic stroke patients. This was a prospective, randomized controlled study. Twelve patients with chronic stroke (>24 months) who were under grade 3 in dorsiflexor strength with manual muscle test were included and randomized into intervention (Dynamic FES) ($n_1$=7) and control (FES) ($n_2$=5). Both the Dynamic FES group and FES group were given a neuromuscular development treatment. The Dynamic FES group has implemented a total of 60 minutes of exercise treatment and gait training with Dynamic FES application. The FES group, with the addition of applying FES while sitting, has also implemented a total of 90 minutes of gait training on treadmill after the exercise treatment. Both two groups accomplished the program, twice a week, for a total of 24 times in a 12-week period. Exercise treatment, gait training on treadmill, and both Dynamic FES and FES were implemented for 30 minutes each. Korean version activities-specific balance confidence scale (K-ABC) was measured to determine self-efficacy in balance function. Timed up and go (TUG) test was performed to evaluate the physical performance. K-ABC, TUG, Berg balance scale (BBS), modified physical performance test (mPPT) and G-walk were evaluated at baseline and at 12 weeks. After 12 weeks, statistically significant differences (p<.05) were apparent in the Dynamic FES group in the changes in K-ABC and BBS. mPPT, TUG, gait speed, stride length and stance phase duration (%) were compared with the FES group. K-ABC had higher correlation to BBS, along with mPPT to TUG. Our results suggest that walking with Dynamic FES in chronic stroke patients may be beneficial for improving their balance confidence, functional ability and gait.