Heating magnetic nanoparticles (MNPs) with alternating magnetic fields (AMFs) have applications in biomedical research and cancer therapy. Accurate measurement of the heating efficiency or specific loss power (SLP) generated by the MNPs is essential to assess response(s) in biological systems. Efforts to develop standardized equipment and to harmonize results obtained from various MNP samples and AMF systems have met with little success. Without a standardized magnetic nanoparticle or calorimeter device, objective comparisons of estimated thermal output among laboratories remain a challenge. In addition, the most widely used adiabatic initial slope model fails to account for thermal losses, which are unavoidable. We propose a non-adiabatic method to analyze MNP heating efficiency derived from the Box-Lucas equation, wherein the sample is subjected to several short duration heating pulses. SLP is then estimated from an arithmetic average of the Box-Lucas fitted coefficients obtained from each pulse. Heating experiments were conducted with two identical samples that were placed within vessels having different thermal insulation using the same AMF parameters. Though the samples generated different temperature curves, the pulsed Box-Lucas method produced nearly equivalent SLP estimates. Further, the pulsed test enabled analysis of the heat transfer coefficient providing quantitative measures of sample heat loss throughout the test, with robust statistical confidence. We anticipate this new methodology will aid efforts to standardize measurements of MNP heating efficiency, enabling direct comparison among varied systems.