This study has been carried out to observe the driving and the intercepting effects of air screen on some sea-water fishes ; Chrysophrys major, Mylio macrocephalus, Fugu pardalis. 1. By perforating fine holes on the compressed air pipes, air screen was formed in the water, and this screen was set at $45^{circ}$, and $90^{circ}$ against the fish passage to observe the reactions on the part of the fish. 2. Changes were given to the location of the air screen in the water to observe how the air screen drived the fish to its direction. 3. The effective distance between holes on the air piprs was determined by a series of experiments of setting up two air screens of various types in hole distance, by moving a screen to-ward the other, and of observing the distance between two screens when fish eseape through the space. The results of the above experiments observed as follows were : 1. The passage of fish was effectively intercepted by setting up the air screens at 45 degrees against the fish passage and it was also intercepted when the screen was set at 90 degrees against the passage. 2. Fish could be driven by moving the air screen toward the fish. 3. The air screen formed from the pipe above than 0.3mm in diameter was effective, but less than 0.2mm was not sufficiently effective. 4. The strength of the air pressure in the pipe should be higher than $0.087kg/cm^2$. 5. The fish holding ability to obtain effective air screen was ranged as following when the air pressure was $0.160kg/cm^2$. and the hole diameter was 0.3mm on the 12.6mm pipe, depending upon the intervals of the holes on the pipe: The shortest distance which could hold fish between two screens was 59.4cm when the holes were perforated at every 40cm; 33.5cm when at every 30cm; 28.75cm when at every 10cm: and 27.25cm when at every 5cm. Thus, no significant change was at served when the holes were perforated more densely than 30cm intervals. Therefor the hole intervals should be 30cm in designing fishing gear employing air screen.