Abstract: This dissertation explores ultrasound imaging of teeth and their internal structures. Of particular interest are those structures that are difficult to detect with radiography and other imaging modalities, including microcracks, subsurface caries and fractures, occlusal caries, and defects that are adjacent to restorations. Ultrasound is well suited for this application due to its high resolution, its ability to detect small variations in acoustic impedance, its ability to penetrate hard tissues and restorations, and its non-ionizing nature. A commercially available water-coupled PZT hydrophone was first used for preliminary ultrasound studies on teeth, and provided a complete circumferential image of the enamel surface and the underlying dentino-enamel junction (DEJ) of a human molar. Subsequently, an imaging system was designed, fabricated, characterized, and tested specifically for the challenging task of imaging high acoustic impedance hard dental tissues. This imaging system was comprised of an 18.9 MHz single element PLZT transducer, a 17.4 MRayl gallium-indium alloy acoustic couplant, a custom-made transceiver, a sophisticated mechanical setup, and a number of tooth phantoms that were created to simulate healthy teeth as well as various irregularities, such as fractures, carious lesions, and restorations. Due to the improved operational efficiency of the system (-31 dB), the narrow acoustic beam (8° FWHM), and the increased signal-to-noise ratio, the PLZT system was able to provide superior images of the enamel and DEJ surfaces in a human molar. Using various tooth phantoms, the imaging system was also able to demonstrate the ability to detect a 25 μm dentinal fracture, a subsurface carious lesion, and the penetration of resin-based composite, porcelain, silver-mercury amalgam, and gold restorations. These results indicate that the intended long-term objective of developing a handheld ultrasound scanner for clinical use as a complement to dental radiography is within reach.