The reliability of various functional coatings (tribological, corrosion-resistant, thermal-barrier, magnetic, electrical, paints) is related directly to the integrity of the interface, which in turn depends upon the occurrence of such events as the nucleation and growth of microflaws under operating conditions. The same is true for the bond between a composite repair patch and a metal substrate as widely used in aging aircraft. For all these applications, it is of interest to detect deteriorated areas of the interface much before the occurrence of large area peel-offs between maintenance cycles. The only possibility of interrogating the quality of such bonds using NDE methods is to detect interfacial flaws in the range of 10 to 100 µm in lengths, much before they grow into fully developed delaminations. Detection of these microflaws is accomplished in this paper by using a recently developed and patented technology for generating well-defined stress pulses with nanosecond (10-9 sec) rise-times and a 15 to 20 nanosecond (ns) duration. A 2.5 ns long specially-configured laser pulse is made to impinge on the back surface of a substrate disc, with a test coating of interest deposited on its front surface. Upon laser/material interaction, a compressive stress pulse of 1 ns rise-time is generated in the substrate disc and propagates towards the test coating/substrate interface. the stress pulse transmits through the flawed interface, and arrives distorted at the coating's free surface, carrying the history of its interaction with the interfacial flaw(s). The profile of this distorted stress pulse is recorded by measuring the resulting transient displacement history of the coating's free surface by using a state-of-the-art optical interferometer which is capable of recording fringes from even optically rough surfaces and with a resolution of only 0.2 ns in a single short mode. The above strategy is also used to record the input stress pulse from an uncoated substrate surface. Both the distorted and parent stress pulse profiles are used as an input to the inversion routines, giving both he size and shape of the interfacial flaw. also, by changing the interrogating area, it is possible to detect both single as well as populations of interfacial flaws. In order to promote further development of the inversion techniques so that they converge faster and more accurately to actual interface flaw geometries, experiments were conducted on samples with well-characterized flaw geometries. In these experiments, interfacial flaws of 10 to 100 µm in diameter were artificially generated in a Nb/sapphire interface system by using the photolithographic technique. In this strategy, a test coating deposited on the substrate is forced to buckle under the influence of a residually-strained Nb top layer. The buckling sites are controlled through an intervening layer of photoresist that is patterned with circular holes with dimensions approximately equal to that of the desired crack size. Because of the highly complaint photoresist layer, the residual stress from the Nb loading layer is fully relaxed and is not transmitted to the lower test layer. However, areas devoid of the photoresist results in the buckling of the test layer as the residual compressive stress of the Nb coating is fully transmitted to the lower test layer. After buckling, the photoresist and the Nb loading layers are removed by dissolving a sacrificial Al layer in HCl. This leaves a well-characterized interface flaw distribution with prescribed size and profile of individual cracks. Both circular and line flaws with dimensions anywhere between 10 and 500 µm could be generated using this strategy. One such well-characterized interface flaw was loaded with the parent stress pulse, and the distorted stress pulse arriving at the coating's free surface was recorded by using the optical interferometer. The information on interface flaw geometry and density provided the additional information for fine-tuning and further development of the numerical surface inversion routines. The paper will discuss the application of the above procedure to several metal/ceramic interfaces of current engineering interest.