ABSTRACTThe ability to predict thermal diffusivity from heat penetration data containing normally distributed errors in can dimensions, measurements in time, measurements of temperature, thermocouple probe location, and assumptions concerning surface heat transfer coefficients was investigated using a nonlinear least squares solution of Fourier's heat conduction equation. Thermal diffusivity calculated from heat penetration data is largely dependent on errors associated with temperature measurement and to a lesser extent dependent on errors in thermocouple probe location. Errors in can dimensions and time measurement had only a minor influence on the prediction of thermal diffusivity. Best prediction accuracy is obtained under the following conditions: (1) When a large can with a length over diameter ratio close to 0.8 is used; (2) when the difference between the initial and heating medium temperature is greater than 40°C; (3) when the data used for calculation of thermal diffusivity is limited to the temperature ratio range of 0.15 to 0.85; (4) when the Biot Numbers for the surfaces of the can are measured or known to be greater than 200; (5) when an accurate time clock is used, and when the position of the thermocouple probe is accurately known