This paper illustrated an optimization method of sensor layout in air duct system, aiming at detecting toxicant before any catastrophic consequence in a bio-terrorist attack. The method was based on genetic algorithm (GA) with minimal emissions as objective function of the optimization procedure. In this paper, we discussed the impact of the number of sensors, the distance between nodes and the accuracy of sensors on the optimizing objective function. Results of a case study showed that the minimal emission of contaminants decreased significantly as more sensors were set in the ventilation system. However, the decrement provided by every sensor was only about 2% when the number of sensors exceeded three. Thus, the optimized number of sensors in this case was supposed to be three, considering the high cost of each sensor. Moreover, the value of the objective function was reduced by 59.3% with sensor layout optimization applied with three sensors while the effects of the accuracy of sensors and the distance between nodes were not significant. In conclusion, adopting the sensor layout optimized with GA ensured the detection of contaminants with the minimal distribution into rooms, which enabled a security control center to learn the concentration and location of the contaminants and respond quickly.