Experiments were performed to determine the noise characteristics of a hydrophone streamer that had incorporated a number of noise reduction features. In the original system, the channels to which the depth-controller birds were attached were 3 to 4 times noisier than nonbird channels. Fortunately, the bird noise is near-field and is eliminated simply by increasing bird/hydrophone separation to 9 ft. On this cable, no other discrete noise sources are evident. The boat, propulsion system, lead-in cable, tail buoy, and ambient sea conditions (moderate seas) do not generate significant noise at towing speeds above 5 knots.The noise on individual hydrophones not near birds is mainly random with only a small coherent component traveling horizontally through the water from the direction of the boat. However, since the 145-ft hydrophone arrays of 20 detectors are much more effective in reducing random noise than coherent noise, the array output consists of approximately equal portions of each. A twofold decrease in the total noise-to-signal ratio would result from doubling the array length (to 290 ft) while maintaining the same hydrophone density. This would result in a four to fivefold decrease in the coherent noise-to-signal ratio and a 30 percent decrease in the random noise-to-signal ratio. Additional noise reduction would result from increasing the hydrophone density and decreasing the motion sensitivity of the hydrophones. (The streamer hydrophones are not the motion canceling type).At a towing speed of 5.3 knots, the noise level recorded on an array (not near a bird) is equivalent to pressures of 1 mu bar. In normal operations with an 8-gun sleeve exploder source, a stacked section signal-to-towing noise ratio of 3 was obtained at 3.0 sec. However, the towing noise increases as the cube of the boat speed, and the S/N ratio would decrease by a factor of 11 if the boat speed were doubled. Conversely, decreasing the boat speed by 18 percent would double the signal-to-towing noise ratio.