A method for identifying an unknown gas using a diffusion limited sensor providing an electrical signal output when exposed to the gas. According to the method, a sensor is disposed in a chamber of fixed volume and having a gas input and gas output through which the gas flows, and the chamber is connected to a source of a known test gas of known concentration and known diffusion coefficient D.sub.k. The input to the chamber is opened, causing the sensor to be exposed to the known test gas until a steady state electrical signal output from the sensor is obtained. The time T.sub.k necessary for the sensor to attain a predetermined percentage of the steady state output is determined. The chamber is then connected to a source of unknown gas, and the input is opened, and time T.sub.u necessary for the sensor to attain the predetermined percentage of a steady state output for the unknown gas is determined. The diffusion coefficient D.sub.u for the unknown gas is calculated by the formula D.sub.u =D.sub.k.times.(T.sub.k /T.sub.u); and the unknown gas is identified by its calculated diffusion coefficient.