In contrast with thermogravimetric analyzer (TGA), where the samples stacked in the crucible inhibit the release of primary pyrolysis products that subsequently undergo secondary pyrolysis, and fluidized beds which fail to provide precise data on the yields of pyrolysis products without secondary reactions, the wire-mesh reactor (WMR) effectively avoids secondary pyrolysis and provides more accurate pyrolysis data because of the discrete and loose single-layer posput of sample particles. Meanwhile, TGA testing is far from standard practice because of its limited heating rate. Thus, the pyrolysis characteristics of four biomasses are comparatively studied in TGA with heating rates of 5, 10, 20, and 40 degrees C.min(-1) and a homemade WMR with heating rates of 5, 40, 250, and 1000 degrees C.s(-1). Both TGA and WMR experiments demonstrate that increasing the heating rate shifts the main pyrolysis curve to higher temperatures due to the thermal hysteresis effect; however, the biomass pyrolysis yield is higher in WMR than in TGA. On average, the activation energies from WMR were higher than from TGA (approximately 14 and 6 kJ.mol(-1) for herb biomasses and woody biomasses, respectively). WMR experiments indicate that it becomes diffusion controlled at 500 C though biomass pyrolysis essentially complete at 500 C; therefore, a further increase in pyrolysis temperature does not exhibit an increase in the product yield. The heating rate exhibits an opposite effect on the yields of tar and gas; meanwhile, both herb biomass and woody biomass show reversed results. The biomass with high H/C and O/C, as well as low H/O molar ratios, yields relatively low amounts of tar and total volatile, as well as high gas, and vice versa. (C) 2017 Elsevier Ltd. All rights reserved.