We aimed to investigate effects of optic-flow direction and visual jitter on the vection during walking. In Experiment1, viewpoint motion along line-of-sight was simulated in a three-dimensional cloud of dots (4km/h), and its optic flow was presented on a 120-inch rear-screen. The direction of simulated motion was forward (expansion) or backward (contraction). Visual jitter (vertical oscillation at 1Hz, amplitude 10cm) was added on the optic flow for half trials. Eleven subjects observed the stimulus during walking (4km/h) or standing still on a treadmill at 1.2m viewing distance for 60s. Vection latency was measured. All 8 conditions (forward/backward x with jitter/without jitter x walking/standing) were repeated 4 times. We found the near-significant interaction of walking and optic-flow direction conditions (p=.08). The latency was shorter with the backward optic flow than the forward only when observers were walking. The interaction of walking and jitter conditions were significant (p<.05). The jitter facilitated vection only when observers were standing still. In Experiment2, we simulated viewpoint motion horizontally (leftward/rightward). Any main effect or interaction was found. These results suggest the vection is inhibited when the self-motion directions in vision and action are identical, and the visual-jitter effect is limited for stationary observers.