Ultrasound-guided needle interventions are common procedures in medicine, and tissue-mimicking phantoms are widely used for simulation training to bridge the gap between theory and clinical practice in a controlled environment. This review assesses tissue-mimicking materials from 24 studies as candidates for a high-fidelity ultrasound phantom, including methods for evaluating relevant acoustic and mechanical properties and to what extent the reported materials mimic the superficial layers of biological tissue. Speed of sound, acoustic attenuation, Young's modulus, hardness, needle interaction forces, training efficiency and material limitations were systematically evaluated. Although gelatin and agar have the closest acoustic values to tissue, mechanical properties are limited, and strict storage protocols must be employed to counteract dehydration and microbial growth. Polyvinyl chloride (PVC) has superior mechanical properties and is a suitable alternative if durability is desired and some ultrasound realism to human tissue may be sacrificed. Polyvinyl alcohol (PVA), while also requiring hydration, performs well across all categories. Furthermore, we propose a framework for the evaluation of future ultrasound-guided needle intervention tissue phantoms to increase the fidelity of training programs and thereby improve clinical performance.