A lab-scale two-stage reactor has been constructed for studying the release and destruction of tars in the two-stage gasifier. First, the pyrolysis characteristics of three fuel samples are investigated only using the single stage reactor. The results show that the maximum value of tar yield is: rice straw 25%, corn straw 22%, and fir sawdust 31% of the initial fuel. Then, the experimental program is extended to investigate the effect of operating conditions in the second stage of the reactor on tar removal. The effects of temperature, residence time, char particle size, char type, fuel type, and diluted air feeding to throat on tar emission has been studied. The results show that the tar decreased with increasing temperature and residence time and with decreasing char particle size. The char type has little effect on tar reduction. Tar emission with limited diluted air feeding is obviously less than that with empty second stage due to the more reactive radicals produced in oxidative conditions. The straw tars appear to have a different suite of compounds than the other two samples of derived material and presumably have different cracking pathways. The tars collected from first stage and second stage have been characterized by gas chromatography/mass spectrometry (GC/MS) and gel permeation chromatography (GPC). The results indicate that tar after pyrolysis contains a large amount of oxygenated constituents. With the increasing of reaction severity (from the empty heated second stage to heated second stage with char bed), the tar compounds reacted further (polymerized) to form larger molecular mass material. It is clear that the material characterized by GC/MS represents a very small part of the total tar. The results have shown that the tar emission from two-stage gasifier can reduce to low levels using optimized operating conditions, but complete tar removal is difficult to realize due to manipulation of operating parameters and fuel type.