The advancement of additive manufacturing methods for the production of metallic parts has initiated the potential development of materials with tailored microstructures to enhance their material properties. To help facilitate the development, methods based on ultrasonic grain scattering are proposed to provide in-situ monitoring of the microstructure’s evolution during the build process. In this work, the longitudinal attenuation coefficient is considered, theoretically and experimentally, as a function of temperature during an annealing process of steel. Theoretically, an iterative solution to the attenuation model of Stanke and Kino is given. The theory is compared against experimental measurements of the longitudinal attenuation coefficient for a steel sample taken at various stages of annealing. Laser ultrasound was employed because it is a remote technique that minimizes unwanted temperature related effects. The annealing process brought the sample from room temperature to 950 o C. A phase transformation from ferrite to austenite occurred at 800 o C, which caused a significant drop in the measured attenuation coefficient. The theoretical attenuation model borrowed previously measured temperature-dependent single-crystal elastic constants of pure iron as model inputs. A mixing formula that considers the volume fraction of ferrite to austenite was applied near the 800 o C mark where the drop in attenuation appeared. Remarkably, the theoretical attenuation model almost exactly reproduced the experimental data points. This concurrence supports: (1) the employment of laser ultrasound for measurement of the attenuation during heating of materials, (2) the suitability of theoretical ultrasonic grain scattering models during highly transient temperature behavior, and (3) the ability of the theoretical attenuation model to represent the effect of a phase transformation.