I propose to study the coordination mechanisms used by teams of distributed volunteers involved in Free, Libre and Open Source Software (FLOSS) development, which has been successful despite the dual challenges of computer mediated distributed work and volunteer management. There is an opportunity to understand how the motivations of participants affect the manner in which they organize their work and to learn from their success. I propose to conduct a case study of a FLOSS project that has collaborated in each of four modes made of volunteer/nonvolunteer and distributed/collocated axes. Drawing on archives and interviews, I will use coordination theory process mapping to describe the coordination mechanisms and sequence analysis to test for differences between the four settings. Participant observation and a literature review reveal examples of novel coordination mechanisms that seem especially associated with distributed volunteers. These are dependency minimization and ‘post-hoc’ coordination, where the group acts to integrate individually chosen contributions. A possible explanation of the use of such mechanisms is that partners expect each other to be unreliable and have developed mechanisms to facilitate effective collaboration in the face of that challenge. It is hoped that these novel mechanisms will be useful to those seeking to draw on unreliable collaboration partners in an effective and scalable manner, including research and development teams, some scientific collaborations and networks of small firms collaborating without formal contract arrangements.