An Uzi submachine gun, held in a machine rest, was fired into a large container of water at various controlled angles. Initially, the weapon was fired in a horizontal position 30 times. Bullet flight stability was checked with verification paper at 0.50 and 4.75 meters. Velocity was measured at 2 meters, and average velocity and standard deviation for the lot of ammunition were determined. A minimum of 2 minutes was allowed between shots. The roundness of holes in the paper was examined individually for each shot, and the paper was changed after each shot. Study of the ricochet was started at 2 degrees, and the angle was then gradually increased until the critical angle at which the bullet no longer ricocheted was reached. At each angle and with each shot, a piece of computer paper was floated on the entire surface of the water to determine the point of impact and the track of the bullet while it remained in contact with the water. Angles of incidence and departure relative to the water were calculated by measuring horizontal and vertical distances from the muzzle to the point of impact on the water and from the point of departure from the water to the verification paper behind the water tank. Experimental results showed that the measured critical angle of ricochet for a 9-mm parabellum bullet fired from the Uzi was 6.5 degrees. The maximum depth of penetration into the water during ricochet was 22 mm, and the maximum track length in the water was 62 cm. It is concluded that a relatively shallow pool of water can ricochet a bullet without the bullet ever coming in contact with the underlying surface. At a weapon height of 1 meter, about hip level for a man, a fired bullet can strike water at a distance as close as 8.8 meters and remain unmarked if the depth is a mere 2.5 cm. 3 references, 1 table, and 8 figures (Author abstract modified)