No standard method has been developed for measuring the evolution of specific toxic gases from building linings when involved in fire. The British Fire Propagation test (BS 476 Part 6) operated in an instrumented room has been proposed for this purpose previously but has not found general acceptance. It is considered further in this report, which investigates ihe movement and measurement of smoke and specific fire gases under different conditions of room stirring and the effect of the latter on fire propagation indexes. Stirring has been found to have no statistically significant effect on fire propagation indexes provided that the effects of this on calibration of the apparatus are taken into account. Stirring also had little effect upon smoke production per se. Under unstirred conditions smoke and toxic gases stratify in the same layer early in the test, and measurement of their production a t any single room location will be subject to the location, the way the room influences stratification and how the room is instrumented, as well as by the product performance. Under stirred room conditions smoke and toxic gases are evenly distributed and product performance can be assessed more simply from concurrent measurements of fire, smoke and toxic gas parameters. The latter procedure is proposed for obtaining relative data on building linings and for examination in further studies for correlation to room and corridor burns.