Convergence of the eyes, that is, the tendency for the eyes to independently point towards an object of interest, is often listed as a depth perception cue. It's true that convergence is a part of depth perception, but in fact, it is more of an effect than it is a cause. The real cause is binocular disparity, or parallax.
Objects in three dimensions, when viewed independently from two distinct vantage points, appear to be at slightly different angles (unless the object is colinear with the two vantage points, but depth perception colinear with the eyes is impossible - the eyeballs would have to be transparent!). The greatest angular resolution of the eyeball is near the centre-line, so when a person is looking at an object, to get the most information regarding the angular disparity, each eyeball rotates so as to project the object to the centre of the retina. This increases the accuracy of angular measurement, and hence increases the accuracy of depth perception. But there would be no convergence if there were not disparity to prompt the eyes to converge, and even with convergence it is the disparity (coupled with the intrinsic knowledge of the difference in angle between the eyeballs) that is used to estimate depth.