Advanced piezoelectric-based ultrasonic transducers offer the potential for in-coolant nondestructive testing (NDT) measurements at high temperatures (HTs), including during hot standby (~260°C) for liquid-sodium–cooled advanced small modular reactors. The reliability of the NDT measurements is typically quantified by the probability of detection (POD) measured at the corresponding temperature. Obtaining such data in liquid sodium is challenging. Using a model-assisted POD approach, a transfer function is reported that enables data obtained on low carbon steel specimens at room temperature to give an estimated POD at an HT. A primary source of the difference in POD between room temperature and HT is due to the transducer material temperature-dependent performance. This paper demonstrates the transfer function approach using data for modified lead zirconium titanate (PZT-5A). A physics-based model was developed using a finite element method and used to quantify reduction in the scattering amplitude for standard reflectors, side drilled holes (SDHs), for a range of sizes, from 15°C to 195°C. Scattering amplitudes for the room-temperature–simulated data are compared with the experimental data measured at 2.25 MHz. A temperature correction and transfer functions were developed to transform the simulated temperature effect in the physics-based model to compare with the experimental data. The model-based approach was validated with experimental data. It was seen and validated for a PZT-5A ultrasonic transducer operating at 2.25 MHz that the 95% POD at 15°C was 0.58 λ, and due to variation in temperature-dependent properties of PZT-5A, the 95% POD was achieved only for a 1.41 λ SDH diameter.