Aims/hypothesis Increasing brown adipose tissue (BAT) activity is a possible therapeutic strategy to increase energy expenditure and glucose and lipid clearance to ameliorate obesity and associated comorbidities. The thiazolidinedione (TZD) class of glucose-lowering drugs increase BAT browning in preclinical experimental models but whether these actions extend to humans in vivo is unknown. The aim of this study was to determine the effect of pioglitazone treatment on adipocyte browning and adaptive thermogenesis in humans. Methods We first examined whether pioglitazone treatment of cultured human primary subacromioclavicular-derived adipocytes induced browning. Then, in a blinded, placebo-controlled, parallel trial, conducted within the Baker Institute clinical research laboratories, 14 lean male participants who were free of cardiometabolic disease were randomised to receive either placebo (lactose; n = 7, age 22 ± 1 years) or pioglitazone (45 mg/day, n = 7, age 21 ± 1 years) for 28 days. Participants were allocated to treatments by Alfred Hospital staff independent from the study via electronic generation of a random number sequence. Researchers conducting trials and analysing data were blind to treatment allocation. The change in cold-stimulated BAT activity, assessed before and after the intervention by [ 18 F]fluorodeoxyglucose uptake via positron emission tomography/computed tomography in upper thoracic and cervical adipose tissue, was the primary outcome measure. Energy expenditure, cardiovascular responses, core temperature, blood metabolites and hormones were measured in response to acute cold exposure along with body composition before and after the intervention. Results Pioglitazone significantly increased in vitro browning and adipogenesis of adipocytes. In the clinical trial, coldElectronic supplementary material The online version of this article (https://doi.org/10.1007/s00125-017-4479-9) contains peer-reviewed but unedited supplementary material, which is available to authorised users. -017-4479-9 induced BAT maximum standardised uptake value was significantly reduced after pioglitazone compared with placebo (−57 ± 6% vs −12 ± 18%, respectively; p < 0.05). BAT total glucose uptake followed a similar but non-significant trend (−50 ± 10% vs −6 ± 24%, respectively; p = 0.097). Pioglitazone increased total and lean body mass compared with placebo (p < 0.05). No other changes between groups were detected. Conclusions/interpretation The disparity in the actions of pioglitazone on BAT between preclinical experimental models and our in vivo human trial highlight the imperative to conduct human proof-of-concept studies as early as possible in BAT research programmes aimed at therapeutic development. Our clinical trial findings suggest that reduced BAT activity may contribute to weight gain associated with pioglitazone and other TZDs.Trial registration ClinicalTrials.gov NCT02236962