Software is present in every aspect of our lives, pushing us inevitably towards a world of distributed computing systems. Agent concepts hold great promise for responding to the new realities of large-scale distributed systems. Multi-agent systems (MASs) and their underlying theories provide a more natural support for ensuring important agent properties, such as autonomy, environment heterogeneity, organization and openness. Nevertheless, a software agent is an inherently more complex abstraction, posing new challenges to software engineering. Without adequate development te- niques and methods, MASs will not be sufficiently dependable, thus making their wide adoption by the industry more difficult. The dependability of a computing system is its ability to deliver a service that can be justifiably trusted. It is a singular time for dependable distributed systems, since the traditional models we use to express the relationships between a computational process and its environment are changing from the standard deterministic types into ones that are more distributed and dynamic. This served as a guiding principle for planning the Software Engineering for Large-Scale Multi-Agent Systems (SELMAS 2006) workshop, starting with selecting the theme, "building dependable multi-agent systems." It acknowledges our belief in the increasingly vital role dependability plays as an essential element of MAS development.