Purpose: To characterize brown adipose tissue (BAT) in polycystic ovary syndrome (PCOS) patients in comparison to healthy subjects using Z-spectral imaging (ZSI). Method: ZSI data were collected on 19 normal control females (NCF), 17 males (NCM), and 13 PCOS patients. By fitting to multiple Lorentzian functions, ZSI provides fat-water fraction (FWF) of tissue in the supraclavicular area that can be used to differentiate between white adipose tissue (WAT), BAT, and muscle. The fraction of BAT over the total fat depot (BATf) and the average FWF in BAT or FWF(BAT) were then computed, reflecting relative BAT mass and BAT metabolism respectively. The parameters were compared among the three groups, and the correlations to Body Mass Index (BMI) were also quantified. Results: There was an inverse correlation between BATf and BMI in normal subjects. The BATf of the PCOS group was significantly smaller than the NCF (P < 0.001). On the other hand, FWF(BAT) correlated linearly with BMI in healthy subjects. The PCOS group had higher FWF(BAT) than the NCF group (P < 0.001). Conclusions: Normal subjects with higher BMI show less BATf and have increased FWF(BAT), indicating relatively higher level of metabolic passive WAT depot and relatively reduced metabolism in their BAT depots. PCOS patients have the least BATf and the highest FWF(BAT), suggesting decreased BAT mass and function in PCOS. Novel imaging technique with ZSI for the characterization of BAT mass and function in PCOS may help to monitor treatment responses of PCOS therapies.