One of the key elements of precision agriculture is the farm management information system (FMIS) that is responsible for data management, analytics and subsequent decision support. Various FMISs have been developed to support the management of farm businesses. A key artefact in the development of FMISs is the software architecture that defines the gross level structure of the system. The software architecture is important for understanding the system, analysing the design decisions and guiding the further development of the system based on the architecture. To assist in the design of the FMIS architecture, several reference architectures have been provided in the literature. Unfortunately, in practice, it is less trivial to derive the application architecture from these reference architectures. Two underlying reasons for this were identified. First of all, it appears that the proposed reference architectures do not specifically focus on FMIS but have a rather broad scope of the agricultural domain in general. Secondly, the proposed reference architectures do not seem to have followed the proper architecture documentation guidelines as defined in the software architecture community, lack precision, and thus impeding the design of the required application architectures. Presented in this article is a novel reference architecture that is dedicated to the specific FMIS domain, and which is documented using the software architecture documentation guidelines. In addition, the systematic approach for deriving application architectures from the proposed reference architecture is provided. To illustrate the approach, the results of multi-case study research are shown in which the presented reference architecture is used for deriving different FMIS application architectures.