In the last decades, the recognition of an endocrine role for white adipose tissue constituted a breakthrough in the understanding of the cross-talk among tissues for regulation of metabolism and for many other processes, from inflammation to the control of feeding behaviour. In contrast to white adipose tissue, the capacity of brown adipose tissue (BAT) to synthesize and release most of the currently recognized adipokines, is very low. In fact, the capacity to produce adipokines such as adiponectin or even leptin is considered a marker sign of white-versus-brown identity in adipose cells (1). Retinol binding protein-4, a protein with adipokine properties, is released both by white and brown adipose tissue, but there is a specific regulation in brown adipocyte in response to noradrenergic and PPARalpha-mediated signals (2). An exception to the traditional concept of a poor secretory capacity of brown adipocytes is triiodothyronine (T3). Differential expression of 5′-deiodinase in brown but not in white adipocytes makes brown adipose tissue an active site of T3 synthesis by deiodination of T4. In the late 80′ it was shown that locally generated T3 is not only involved in intracellular regulation of brown adipocyte thermogenesis but also makes brown adipose tissue a quantitatively relevant site of production of systemic T3 in conditions of highly recruited brown adipose tissue thermogenesis (3). Other signalling molecules, such as several interleukins and IGF-1, have been claimed to be synthesized and released to circulation by brown adipose tissue (1). We have found that brown adipose tissue is an active site of synthesis and release of the hormonal factor fibroblast growth factor-21 (FGF21). Noradrenergic-mediated thermogenic activation of brown fat induces a cascade of intracellular signalling involving activation of p38-MAP kinase that ultimately results in activation of FGF21 gene transcription through phosphorylation of ATF2 and binding to the FGF21 gene promoter. This results in the induction of FGF21 synthesis and FGF21 release by brown adipocytes. Direct assessment of FGF21 release “in vivo” by determination of arterio-venous differences in FGF21 concentration across interscapular BAT evidenced a significant FGF21 output in conditions of thermogenic activation of BAT (4). Calculations based on the FGF21 turn-over rate and FGF21 total output release by interscapularar BAT, indicate that BAT is a substantial source of systemic FGF21 in conditions of enhanced thermogenesis. Recognition of an endocrine role of BAT for the FGF21 system suggest that some of the beneficial actions of BAT activation on systemic metabolism may be mediated by an active release of FGF21 and/or other yet unidentified hormonal factors that favour glucose homeostasis and a healthy metabolic profile. Moreover, an active endocrine role of BAT may explain the association between BAT activity and systemic metabolism even in conditions in which BAT amounts are scarce (e.g. adult humans) and exclusive attribution of consequences of BAT activity to overall changes in energy balance are unlikely. The combination of multiple experimental approaches, from candidate-based experimentation to high throughput screening, is expected to lead to the identification of the brown adipocyte secretome and its biological significance.