The forward osmosis (FO) process potentially can be used as an alternative to traditional brine treatment technologies due to its low energy requirement. In this study, the feasibility of concentration of brine by the FO process was investigated using a laboratory-scale unit. An FO membrane and the dense selective layers (i.e. with their support layers properly peeled off) of two types of RO membrane (i.e. cellulose acetate asymmetric and polyamide composite, denoted as CA and AD, respectively) were tested. It was found that the dense selective layer of the CA membrane could achieve the highest water flux because it had a relatively high hydrophilicity and was thinner that minimized the occurrence of internal concentration polarization. SEM images of the surfaces of the dense selective layers of both the CA and AD membranes displayed non-porous structures, which contributed to high salt rejections of above 99.7%. Comparing the performance between the FO membrane and the dense selective layer of the CA membrane for concentrating brine, the FO membrane could only achieve an initial water flux of 4.9 gal center dot ft(-2)xd(-1) (GFD; 2.3 mu m/s or 8.2 l center dot m(-2)xh(-1)) while an initial water flux of 8.9 GFD (4.2 mu m/s or 15.0 l.m(-1)xh(-1)) was achieved by the latter at similar experimental conditions. Moreover, water fluxes were kept at a relatively stable level without much declination when the draw concentration was maintained at a consistently high value. The highest recovery achieved by the dense selective layer of the CA membrane after 18 h was about 76%. In addition, white precipitates were found on the wall of the feed container and on the surface of the heater, which indicated the possibility of mineral recovery. This study suggests that the FO process can be potentially used to concentrate brine while further researches are required to optimize the FO process.