A total of 4 tests were conducted, in the 1960's, in a three room arrangement, in a ranch house. The structure was 3.9 m wide and 13.6 m long, and contained three rooms, designated living room, dining room and bedroom, which were all of equal length, in a row, starting with the living room. All windows were closed during the tests, but interior doors were open. All rooms were fully furnished. The exterior walls and roof were fir plywood, and the exterior siding was gypsum board nailed to the “2 x 4” studs. The tests were all initiated by ignition of a 4.4 kg wood crib, in the living room, near an easy chair, which itself ignited very soon afterwards. Measurements made included temperatures, in °F, at 24 locations throughout the structure (including one measuring ambient temperature). Concentrations of three gases were also measured throughout the tests with continuous analyzers: carbon monoxide, carbon dioxide and oxygen. The gas measurements were made in sequence, so that concentrations at each location were measured at different times, but all gases were measured simultaneously at each location. The tests were conducted using 4 different wall lining products: (a) 13 mm (0.5 in) thick painted gypsum board, nailed to the studs; (b) 6 mm (0.25 in) thick pre-finished plywood panels nailed over 13 mm (0.5 in) gypsum board, itself nailed directly to the studs; (c) 16 mm (0.25 in) thick pre-finished plywood panels nailed directly to the studs and (d) 6 mm (0.25 in) thick pre-finished fire-retardant (pressure-treated core) plywood panels nailed directly to the studs. The original data interpretation was restricted to a consideration of whether flashover was or was not reached and little else. A new analysis of the data, however, has permitted determination of broad approximations to heat release (both comparative values and rough average data), and comparisons with data from other standard fire test methods, such as the Steiner tunnel (ASTM E84) and the cone calorimeter (ASTM E1354). This type of analysis is very important because (a) it recovers information which is often lost and (b) it allows significant information to be obtained from existing tests conducted at earlier dates, for semi-qualitative reasons. In this way, cost savings can be achieved in decreasing the need for repeated testing.