Two experiments are reported which attempted to apply techniques derived from an extensive set of psychological theories and research to enhance return compliance in a tuberculosis detection drive. Experiment 1 manipulated four methods of providing return reminders (postcard, telephone call, direct person-to-person or take-home card) in combination with two types of authority sources (expert vs nonexpert). Experiment 2 manipulated types of patient commitment (verbal, verbal plus written or no commitment) and importance-or-returning message (enhanced vs standard) in combination with two types of return reminders (take-home card vs no reminder card). These theoretically-derived procedures failed to significantly enhance return compliance. The experiments are examined for procedural failures and the findings are discussed in terms of the limited applicability of some theoretical notions in health settings and particular populations. It is asserted that through such failures useful knowledge is gained for understanding the complex phenomenon of compliance.