Two semiconductor gas sensors which are sensitive to methanol and acetone were chosen to form a gas sensor array. Response of sensor array combined with pattern recognition technique of artificial neural network was used to carry out the quantitative analysis of gas concentration in a mixture. Separate gas is injected on gas sensor to understand the sensitivity of sensor upon change in gas concentration. Gas mixture of different acetone and methanol gas composition is prepared and response on sensor array is acquired. The acquired data is a function of concentration profile of gas mixture. Data is pre-processed and given to neural network for quantitative analysis of gas concentration of the mixture. The results show that the system can analyze the gas mixture and produce quantitative data with good accuracy.