This dissertation is concerned with the early part of systems design, the purpose of which is that users and their managers can decide which computer-based systems are needed and relevant, and how they can be provided, i.e. purchased, and/or developed, and implemented, in a specific organization. Systems design begins with a recognition of a possible need for computer support in a specific organization, continues until this need has been analysed, described, evaluated, and discussed, and ends with the description of an overall conceptual and functional design for the corresponding computer-based systems. The research objective behind this dissertation aims at developing theories of and approaches to systems design, which should provide a basis for designers to improve their work practices. In other words, this dissertation contributes to a clarification of what is going on in design, and how we, as designers, should deal with the process of designing for a specific organization. The dissertation is based on exploratory empirical studies in one organization using action research as the research approach. The focus of the action research project is on improving work practices for designers by developing methodological guidelines, heuristics, or "rules of thumb". This involves organizational issues comprising of social, political, and managerial/strategical aspects within the organization. I have carried out action research over a period of 1 1/2 years in a public organization in Northern Europe, where I conducted three separate but interrelated design projects. Having a design approach inspired by and comprising of participatory design, ethnographically inspired, and systems approaches, I have done in depth analysis of current work practices, carried out unstructured interviews, observation, video recording, document analysis, and the like. Also I have set up participatory analysis and design workshops. One of the design projects was an experiment with a systems approach, Work Analysis, an approach that originates in an ESPRIT-project, FAOR (Functional Analysis of Office Requirements). The goal and purpose of the dissertation are to contribute to current discussions, among researchers and practitioners, regarding how to do reliable systems design in small complex organizations - small in the sense that the design projects were conducted in an organization with approximately 50 employees, and complex in the sense that the employees had very different roles and relationships. The results in the dissertation are presented in a way where they can, hopefully, initiate and stimulate discussions among other researchers and engaged practitioners within the field of design. I present a number of examples from my empirical studies, which are based on my experiences using various participative techniques, ethnographically inspired approaches, and the conceptual framework and guidelines provided by Work Analysis.