Intra-organizational Complexity and ComputationOrganizations are complex systems. They are also information processing systems comprised of a large number of agents such as human beings. Combining these perspectives and recognizing the essential non-linear dynamics that are at work leads to the standard non-linear multi-agent system conclusions such as: history matters, organizational behavior and form is path dependent, complex behavior emerges from individual interaction, and change is inevitable.Such a view while descriptive, is still far from the level of specificity and predictive richness that is necessary for organizational theory. To increase the specificity and value of our theories we will need to take into account more of the actual attributes of tasks, resources, knowledge and human cognition. In doing so, it will be possible to achieve a more adequate description of organizations as complex computational systems. More importantly, we will also achieve a greater ability to theorize about the complexity of organizational behavior.Intra-organizational computation and complexity is concerned with discovering, modeling, theorizing, and analyzing the fundamental nature of organizations as complex adaptive systems composed of intelligent but constrained adaptive agents. Within computational organization science researchers search for fundamental organizational objects and the mathematical formalism with which to describe their behavior and interactions. In physics, researchers search for laws governing gravitational, electromagnetic, and other fields of force. In both cases, the aim is to discover the most reasonable basis from which, at least in principle, theories of all other processes and behaviors can be derived. In a complex process there are typically many interacting objects (e.g. people or procedures in an organization or particles in physics) and it is rarely possible to proceed to a complete mathematical solution. Systems in which there are complex processes often exhibit non-linear behavior, phase changes in behavior, and often reach dramatically different end states given only minor changes in initial conditions. Computational analysis, e.g., simulation or enumeration, can be used to track and analyze the detailed behavior within and among these objects (people or particles). Whether we are modeling the behavior of people, robots, organizations or atoms -computer modeling at the quantum level becomes extremely complicated as soon as more than a few of these objects are involved. Computational complexity increases and the length of time for the system to be "solved" or "simulated" on the computer increases.Such work is carried out via formal methods -mathematical and computational reasoning. This paper describes complexity theory and computational organization theory. Then a description of organizations as complex computational systems is presented. Specific attention is paid to the role of knowledge management, network theory, computational theory, and the study of the impacts of information an...