Investigations into optimisation of the Perkin-Elmer flame photometric detector (FPD) response have shown that the S-mode is greatly dependent upon nitrogen (low in a manner that is not independent of the oxygen (or air) flow to the detector. On the other hand, helium does not exert as great an influence upon response, although characteristics do follow similar lines to that of nitrogen. In the P-mode, nitrogen flow has little bearing upon the resultant emission response, the O/H ratio being of primary concern. In studies of flow rate changes for different carrier gases using the S-mode, hydrogen proves to be most suitable in preservation of the detector’s intrinsic square law response, whereas nitrogen has a destructive interference; helium only shows this behaviour at very high flow rates. Permeation tubes are used to linearise response and thereby provide information on the exponential response factor by monitoring the changes in recorded peak widths. They also enable the detection limit to be improved four- to fivefold, to 3×10-11 g S.sec-1 for a simple organosulphur compound.