More will be learned about health programs and the implementation of health policy in this country if we pay more attention to issues of program implementation. Of particular use would be more studies which explicitly link program implementation with program outcomes and which recognize the need to combine quantitative and qualitative analysis of program implementation; the use of triangulated methods in focusing on the relationship between program implementation and program outcomes; the incorporation and study of planned variation in the methods of implementing programs; recognition that the process is essentially one of organizational change and innovation, and the incorporation of existing theory and evidence relevant to these issues; and recognition that the ongoing nature of the implementation process requires longitudinal study designs for implementation as well as for outcome assessment. Cronbach [9] has remarked that evaluation research "lights a candle in the darkness, but it never brings dazzling clarity." It may be that more attention to program implementation and better research on the process, such as that suggested in this note, will provide a little more light and will bring if not dazzling , at least modest, improvements in clarity.