Regulatory T cells (T reg cells) have attracted much interest from both basic and clinical immunologists. Although questions remain about their fundamental biology and their clinical potential has yet to be fully realized, considerable advances have been made in the understanding of the differentiation, homeostasis and function of T reg cells. This new knowledge has led to a substantial increase in the number of Foxp3+ T reg cell subpopulations described in the literature and consequently to an increase in the use of new abbreviations and terminology. Furthermore, as the understanding of T reg cell biology has grown, so too has the realization that some aspects of the original terminology are no longer accurate, and its use has become less stringent. At the Third International Conference on Regulatory T Cells and Th Subsets and Clinical Application in Human Diseases held in Shanghai, China, on 13–16 October 2012, a small workshop was convened to discuss T reg cell nomenclature and to develop several recommendations. We hasten to add that this is simply a list of recommendations, and it remains the prerogative of journals to develop their own editorial preferences and for the authors to use the nomenclature they feel best suits their manuscript. However, we support the recommendations noted below.

There were three general issues that provided the momentum for this workshop. First, the terms used for the principal Foxp3+ T reg cell populations—those that differentiate in the thymus, those that differentiate in the periphery and those generated in vitro—are not ideal, as these are, to some extent, inaccurate, ambiguous and/or uninformative. For example, the widely used term'natural T reg cell'is misleading and ambiguous, as it indicates that all other Foxp3+ T reg cell populations are'unnatural'. It also does not convey any useful or accurate information. One feature of these Foxp3+ T reg cell populations that is more informative is the anatomical location of their differentiation. Instead, use of the terms' thymus','periphery'and'in vitro'provides a clear indication of whence the Foxp3+ T reg cells in question are derived. Thus, we would recommend that'natural Foxp3+ T reg cells' instead be referred to as' thymus-derived T reg cells'(Box 1). Consistent with the rationale above, we would recommend that Foxp3+ T reg cells that differentiate in the periphery be referred to as' peripherally derived T reg cells' rather than'induced or adaptive T reg cells'(Box 1). In this context, we noted that terms used to define a T reg cell subpopulation, such as' induced T reg cells', are often used when the location of their differentiation is unclear. Thus, we would suggest that the newly recommended terms' thymus-derived T reg cells' and'peripherally derived T reg cells' be used only when the anatomical location of their differentiation has been clearly demonstrated. When the origin of the T reg cell being studied is unclear, the general term'Foxp3+ T reg cell'would be more appropriate. Finally, to clearly distinguish between Foxp3+ T reg cell populations that are generated in vivo versus those generated in vitro, we would suggest that the term'in vitro–induced T reg cells' be used for all Foxp3+ T reg cell populations generated ex vivo, such as those generated through the use of transforming growth factor-β (Box 1).