The class of low-cost, distributed ubiquitous systems represents a computing mode where a system has small, inexpensive networked processing devices, distributed at all scales throughout business activities and everyday life. The unique features of such a class of ubiquitous systems make the security analysis different from that for the centralized computing paradigms. This paper presents a holistic approach for security requirement analysis for low cost, distributed ubiquitous systems. Rigorous security analysis needs both quantitative and qualitative approaches to produce the holistic view and the robust data regarding the security features that a system must have in order to meet users’ security expectations. Our framework can assist system administrators to specify key security properties for a low-cost, distributed ubiquitous system and to define the specific security requirements for such a system. We applied Bayesian network and stochastic process algebra to incorporate probabilistic analysis to the framework.