The U.S. Forest Service has started a laboratory study with the ultimate objective of determining model laws for fire behavior. The study includes an examination of the effect of such variables as species of wood, density of wood, moisture content, size of fuel particle, spacing, dimensions of fuel bed, wind, and slope on the rate of spread of fire and the partition of energy. Fuel beds in the form of cribs made of square cross-section wood sticks were chosen for diagnostic studies. The crib is ignited at one end and moved in such a manner that the flame is kept in a fixed position in space. The rate of spread of the fire is determined by the rate at which the crib is moved. After an initial period of growth, the fire reaches a steady-state, which permits measurements to be made of certain dependent variables over an extended period. A series of tests were made using five species of wood with varying densities to determine the effect of density on rate of spread. The species were white fir, magnolia, basswood, sugar maple, and longleaf pine. Results show rate of spread decreases with increasing density, and for a given wood density it is practically independent of the species tested. Another series of tests were made using white fir wood at several densities to determine the effect of fuel moisture content on rate of spread. The results show that the effect of moisture content on rate of spread is greater with decreasing density of wood. The temperature distribution measured in the convection zone of test fires was expressed as a functional relationship between two dimensionless groups, one pertaining to atmospheric conditions and convected heat and the other to the height and diameter of the convection column.