An electrically heated bed of silica sand fluidized by air and at 900degreesC has been used to incinerate a wide variety of solids. These included vegetables, meat, plastics, wood, rubber, car-tyres and a Polish coal, as well as two commercially produced waste fuels. Combustion of such solids always involves two, almost sequential, stages: the first is when any moisture and volatile materials are expelled; during the second, the solid remaining from the first stage (i.e. char) is burnt. Some materials float on the hot fluidized sand, occasionally when they are not expected to float; other materials sink, often unexpectedly. Likewise some particles fragment, some produce a cloud of soot, but others do not. Devolatilization times (t(d)) for the bituminous coal varied as M(2/3), the initial mass of the particle raised to the power 2/3. This is attributed to the rate-controlling step being provision of the enthalpy of thermal decomposition by heat conduction from a particle's exterior to a shrinking core of virgin coal. A simple model for such devolatilization is presented. The char's burn-out time also depended on M(2/3), because control is by mass transfer of O(2) from the bed to a particle's exterior. By contrast a particle of fibre fuel (mainly compressed paper) has td approximately proportional to M, but dRDF (densified refuse-derived fuel-actually domestic refuse, extruded after being screened) has td independent of M. The char burn-out times for these two commercial fuels (with large volatile contents) were much less than for the coal and independent of M. Car tyres and rubbers produced large quantities of soot and CO. Plastics produced no char; they evaporated and rapidly produced clouds of hydrocarbon vapour. Such a laboratory-scale fluidized bed, held in a transparent quartz cylinder, is excellent for observing a particle whilst it burns, as well as measuring gaseous emissions, the temperature of a particle whilst it burns and also the rates of devolatilization and char burn-out.