This case-control study examined the influence of alternative performance improvement interventions on compliance with the Integrated Management of Childhood Illness (IMCI) algorithm. The study examined whether the feedback of performance assessment data alone in the use of the algorithm could improve health care worker performance as opposed to feedback plus formal training, and whether quality improvement teams can enhance the impact of performance in specific areas. The study was conducted in Niger from October 1997 to December 1998. Data were collected at 4- and 6-month intervals, using direct observations, interviews with health care workers and mothers and/or caretakers, and physical reviews of health care facilities. Findings suggest that feedback interventions and quality improvement teams can affect compliance with IMCI standards. However, the effects of these interventions were seen in specific problem areas versus across-the-board improvement in compliance. More needs to be done to support health care workers with the IMCI algorithm in the field. Approaches, such as the unified quality assurance/IMCI model and structured supervision tool, are critical in the struggle against childhood mortality in Niger and implementation of IMCI worldwide.