The low efficacy of many of the library and large-scale information systems that have been implemented in the developing countries has been disappointing, and their appropriateness is often questioned in the governmental and educational institutions of more industrialized countries beset by budget-crunching and a very dynamic transformation of information requirements. This article examines the conceptual and practical deficiencies of current misapplications of the systems approach, which result in inappropriate or disappointing systems. It suggests that more importance should be given to the analysis of opportunity costs and contingencies when scarce resources are at stake, and to the consideration of the complex objectives of the institutions that are to benefit from these systems; the choice of information technologies and the economic evaluation of the options must respond to these conditions. It is also suggested that action research principles can help overcome many of the conceptual and practical problems in systems analyses. (Author/KM)