The paper is a transcript of the author's notes of an objective investigation of the thermal, electricity supply and economic problems associated with the warming of surface-type air-raid shelters. It examines—inter alia—the factors of thermal contrast existing between normal buildings and structures of the type under discussion; why, where, when and how heat energy should be transmitted to the occupants; the pros and cons of radiant and convective forms of space-heaters; the thermal capacity of the structure; conditions conducive to surface condensation or sweating, and the like.The design and construction of a floor-warming installation are examined in detail, the performance data obtained during an 8-week test run are analysed, and the paper concludes with a series of estimates of the seasonal consumption of energy, and of the total costs involved in maintaining various standards of thermal comfort in shelters of the type discussed, based on data obtained during the test period.The following conclusions are formulated:—(1) No other form of space-heating provides an equal (or even approachable) standard of thermal comfort for 100% of the occupants of the shelter, in return for the cost of the energy used and the space occupied. Representatives of local authorities, architects and others who have inspected the installation during periods of severe weather have unanimously endorsed this-view.(2) The method provides a satisfactory degree of thermal storage which substantially maintains the specified standard of thermal comfort when the supply of electrical energy is interrupted during periods of peak load (by time switch or otherwise), or by reason of bomb damage to cables, substations, etc.(3) It provides a maximum of safety against burns, shock, etc., due to inadvertent contact with the source of heat energy.(4) The space occupied by the installation is practically nil, as is also the risk of damage by mischievous children and/or ill-disposed persons.(5) The maintenance (excluding the risk of cable failure) is also nil, and an instance of cable failure has still to be experienced.(6) The initial cost of the installation bears a reasonable ratio to the cost of the furnished structure and to the value of the health and morale of its human contents.(7) The total cost of operation compares most favourably with that of the more orthodox forms of space-heating installations, and appears to be substantially lower for an equal standard of thermal comfort.(8) The thermal capacity of substantially constructed buildings of the type discussed (14-in. brick walls and 6-in. concrete floor and roof) is such that it is technically impossible instantaneously—or even rapidly—to induce in them any acceptable standard of thermal comfort for 100% of the occupants, in response to the turn of a switch.(9) This form of space-heater makes no demands on the national stocks of iron and/or steel, and the man-hours of labour per kilowatt capacity involved in its manufacture and installation are probably the lowest for any form of heater.(10) In view of its inherent characteristic of thermal storage at or about room temperature, and its capacity to “carry-on” under conditions of limited-hour supply (during periods of peak-load restriction), it would appear that there is considerable scope for post-war development for certain classes of building.