BACKGROUND: Use of autologous tissue is ideal in breast reconstruction; however, insufficient donor tissue and surgical and donor-site morbidity all limit its use. Tissue engineering could provide replacement tissue, but only if vascularization of large tissue volumes is achievable. The authors sought to upscale their small-animal adipose tissue-engineering models to produce large volumes of tissue in a large animal (i.e., pig). METHODS: Bilateral large-volume (78.5 ml) chambers were inserted subcutaneously in the groin enclosing a fat flap (5 ml) based on the superficial circumflex iliac pedicle for 6 (n = 4), 12 (n = 1), and 22 weeks (n = 2). Right chambers included a poly(L-lactide-co-glycolide) sponge. Other pedicle configurations, including a vascular pedicle alone (n = 2) or in combination with muscle (n = 2) or a free fat graft (n = 2), were investigated in preliminary studies. Serial assessment of tissue growth and vascularization by magnetic resonance imaging was undertaken during growth and correlated with quantitative histomorphometry at chamber removal. RESULTS: All chambers filled with new tissue by 6 weeks, vascularized by the arteriovenous pedicle. In the fat flap chambers, the initial 5 ml of fat expanded to 25.9 ± 2.4, 39.4 ± 3.9, and 56.5 ml (by magnetic resonance imaging) at 6, 12, and 22 weeks, respectively. Adipose tissue volume was maintained up to 22 weeks after chamber removal (n = 2), including one where the specimen was transferred on its pedicle to an adjacent submammary pocket. CONCLUSION: The first clinically relevant volumes of tissue for in situ and remote breast reconstruction have been formed with implications for scaling of existing tissue-engineering models into human trials.