This paper focuses on the development of a test protocol for the analysis of tar collected from an oxy-steam gasification system with a downdraft reactor configuration using biomass. A 10 kg.h(-1) oxy-steam gasification system developed at the Indian Institute of Science, Bangalore was used for the studies. The present work involves qualitative and quantitative analyses to estimate the amount and nature of the tar present in the gas produced. The major focus of the work has been towards establishing protocols to estimate the content and concentration of various species present in the raw and clean gas. The method uses internal standard alongside external standards while analysing compounds using GC-MS and GC-FID. The study clearly establishes the need for using different reference compounds like naphthalene and phenol for the quantification process depending upon the nature of compounds manifesting as tar molecules. Further, it has been established that the conventionally used gravimetric analysis has limitations towards estimating the total amount of tar. GC-MS is used for the identification of the compounds. The results from the study indicate that the average values obtained for clean gas and raw gas are mg.Nm(-3) and 168 mg.Nm(-3) with gravimetric analysis and 37.6 mg.Nm(-3) and 267 mg.Nm(-3) with GC-MS/FID method. The lower hydrocarbons, having the number of carbon atoms in the range of 1-5, which are not considered as tar compounds, are absent in the clean gas and are quantified to be 10.6 mg.Nm(-3) in the raw gas. Finally, the study also captures a technique of analysis and quantification which is of general nature, used generally in organic compound estimation, and therefore can be applied as a generic procedure. The results indicate that the level of tar in the clean gas is low, making the gas suitable for a variety of applications, and the effluent treatment process simpler. The analysis presents an estimate of the tar level obtained with the GC-FID technique using both internal and external calibration curves. These results are compared with results from a similar approach using GC-MS for quantification and from the external calibration curves using GC-FID and GC-MS. These studies confirm that the approach presented here is suitable for the present research activity.