The continuing quest for higher performance levels of modern gas turbine engines has been accompanied by the demand for higher engine operating temperatures. The use of Thermal Barrier Coatings (TBCs) enabled gas turbines to operate at higher temperatures by protecting the blade material (nickel superalloy) while operating in extreme environments. The TBC system typically consists of a bond coat for protection of the nickel–based superalloy against oxidation followed by a top coat consisting of a thermally insulating zirconia-yttria. In addition to the complex gradation in phases, the coatings are subjected to continuous oxidation with service exposure, mechanical loading on rotating parts, fatigue, thermal mis-match and temperature gradients. Hence, the study of failure mechanisms of TBCs become important in deciding operational reliability and service life of the coating. As there are many systems in which the operating temperatures are not high enough to warrant the use of the top coat (ceramic layer), the study of failure mechanisms in superalloys coated with only the bond coat continue to be of great interest. The present work concentrates on the fracture behavior of NiAl bond coats on nickel superalloy and seeks to evaluate the fracture toughness of the coating through the use of micro-machined samples. A review of the relevant literature indicated that while a considerable body of work exists on bulk tests of failure (spalling, splitting, etc.), not much has been reported in the open literature on the evaluation of basic quantities such as the toughness of the coating itself. The present thesis seeks to establish a protocol for the evaluation of toughness and crack propagation mechanisms in coatings through a combination of micro-sample testing that allows fracture to be correlated with location in the film and the use of an analytical model to quantitatively evaluate stress intensity factors in a bi-material system. A system of NiAl coating produced by pack aluminizing is studied for the fracture properties of the coating. Specimen geometries are optimized to enable micro-cracks to be machined and propagated in a low load testing system, such as a depth sensing indenter, so that crack lengths (and position relative to the interface) can be correlated with load. To enable linear elastic theory to be used, dimensions are determined that allow fracture before general yielding. A three point bending test using miniaturized micro-beam specimens of ~ 4 X 0.3 X 0.3 mm is found to be suitable for the above purpose. The technique is a challenging one that requires focused ion beam machining (FIB) along with careful handling and alignment of small samples. The coatings are characterized for their microstructure by electron microscopy to identify compositional variation across the thickness and to determine the thickness of the coating and inter diffusion zone (IDZ). The crack advancement is monitored with increments of loading and the stress intensity factor is evaluated using a program written in “MAPLE” for an edge crack subjected to bending in a bilayered material. Surprisingly, fracture in this system is found to be stable owing to a gradual increase in toughness from the coating surface to the interface. Such an increase from less than 2 to more than 9 MPa m0.5 may be due to the increasing Ni/Al ratio across the thickness of the bond coat. Crack branching is observed as the crack approaches the IDZ and the reasons for such behaviour are not fully understood. This work establishes the viability of this technique to determine fracture properties in highly graded coated systems and may be readily extended to more complex coating architectures and other forms of loading such as cyclic, mixed mode, etc.