The half-open interval procedure seems to offer an inexpensive method of reducing undercoverage in housing unit frames during data collection. Interviewers check the areas near their selected cases, and, if they find any units missing from the frame, give them a chance of selection. However, the effectiveness of the method in the field has not been tested. This paper reviews how the procedure should work and presents evidence from two surveys about its performance in practice. We show experimentally that the procedure often fails to reduce undercoverage and can introduce overcoverage. We conclude with thoughts about the appropriate role for the half-open interval procedure in household surveys in the future.