The random assignment of indivcduals to treatment conditions is often corrupted in field studies. The typical result is nonequivalence between the treatment groups and a corresponding reduc tion in the study's credibilety. The corruption of the assignment process seems to be particularly common in studies involving trickle processing, wherein the research participants appear and are assigned to treatments one at a time over an extended period Several procedures for safeguarding the integrity of random assignment with trickle processing have been proposed. The present article describes a computer-based, alternative method