Tunnel entrance distributions at six bank swallow sand-pit colonies showed consistently nonrandom, too-regular patterns, supporting the hypothesis that the distance between tunnel entrances is determined by territorial disputes at the tunnel mouths. Average nearest neighbor distances, and numbers of burrows per unit area of pit face, were both consistently greater than their random expectations, a paradox that is explained algebraically. Evidence of tunnel coalescence and communal nesting was found in a completely evacuated colony of 30 tunnels. It is argued that bank swallow colonial behaviors evolved to maximize populations on small exposed bank faces along streams and rivers, and that they are still behaving as if only small fractions of large sand pit banks are available. On small areas, regular spacings help to avoid too close tunneling and concomitant bank collapse. It is argued that a tolerance for communal nesting, once coalescence has occurred, involves less tunneling than the forced evacuations of entirely new tunnels by ousted pairs.