Wildfire spread in living vegetation, such as chaparral in Southern California, often causes significant damage to infrastructure and ecosystems. The effects of physical characteristics of fuels and fuel beds on live fuel burning and whether live fuels differ fundamentally from dead woody fuels in their burning characteristics are not well understood. Toward this end, three common chaparral fuels prevalent in southern California, chamise, manzanita, and ceanothus, were investigated by burning them in a cylindrical container. The observed fire behavior included mass loss rate, flame height, and temperature structure above the burning fuel bed. By using successive images of the temperature field, a recently developed thermal particle image velocity (TPIV) algorithm was applied to estimate flow velocities in the vicinity of the flame. A linear regression fit was used to explain the observed time difference between when maximum flame height and maximum mass loss rate occur, as a function of fuel moisture content. Two different methods were used to extract power laws for flame heights of live and dead fuels. It was observed that the parameters defined in the well-known two-fifths power law for flame height as a function of heat release rate were inadequate for live fuels. As the moisture content increases, the heat release rate in the power law needs to be calculated at the time when the maximum flame height is achieved, as opposed to the maximum mass loss rate. Dimensionless parameters were used to express local temperature and velocity structure of live and dead chaparral fuels in the form of a Gaussian profile over different regimes in a fire plume. (c) 2005 The Combustion Institute. Published by Elsevier Inc. All rights reserved.