Two stack loss efficiency measurement methods, the Total Combustible Carbon (TCC) and Combustibles Meter (CM) methods, have been modified for use on masonry heaters. The applicability of the two methods has been verified with 6 tests on two masonry heaters. Each test starts with a cold heater and requires five firings to achieve two different burn rates. The efficiencies calculated for each firing are weighted according to burn rate following EPA Method 28 for wood heaters. The TCC Method uses carbon balances to calculate the chemical energy loss and the dry stack gas mass for calculation of sensible energy loss. The sensible loss that occurs during the off-period, when combustion of wood has stopped, is measured directly by injecting carbon dioxide in the stack and using it as a tracer gas to measure stack flow rate. In both methods the latent energy loss is calculated from wood moisture content and hydrogen content. The CM Method measures losses more directly and is considered the reference method in this work. The chemical energy loss is measured using a flame combustibles meter. The stack flow rate, which is used for the calculation of sensible loss, is measured directly using carbon dioxide tracer gas during both the on and off periods of the appliance. The overall average efficiencies measured by the two methods, in 5 tests on two different appliances, differed by a maximum of 1.7 percentage points of the fuel energy input. On the average they differed by about 1 percentage point. The results of a one-dimensional finite-difference model of the heat exchanger of one of the tested masonry heaters is compared against thermocouple-measured temperatures. For the 3 tests performed the model predicts the measured temperatures to within 12% during the first half of a firing cycle when the burn rate is high, and to within 5% during the second half of the cycle when the burn rate is low. / Master of Science