Blisters are frequently observed in film/substrate material systems, including thermal barrier coatings (TBCs). In this work, new analytical mechanical models are developed to describe, explain and predict the process of blister nucleation, growth and spallation, and redress the limitations of other approaches in the literature. In doing this, a new physical understanding of blister mechanics is acquired. Blisters in film/substrate material systems are sometimes observed to nucleate, grow and spall off, apparently spontaneously, under constant residual stress, for instance, after cooling to room temperature instead of during cooling. The mechanics of this phenomenon is widely considered to be one of the most interesting and challenging instability problems in solid mechanics. Approaches based on buckling have been largely used throughout the past three decades. These approaches generally require an initial interface separation of critical size under a given magnitude of residual stresses for buckling to occur triggering the growth of interface separation. Blister nucleation and growth are, however, frequently observed to occur at sizes that are much smaller than the critical buckling size, which shows the limitation of the buckling-based approaches. This work uses another approach from Wang, Harvey et al. [1, 2], who hypothesised that pockets of energy concentration (PECs) in the form of pockets of dominant tensile stresses on and around the interface drive the nucleation and subsequent development of blisters. According to the hypothesis, PECs provide extra energy in addition to the residual strain energy to nucleate and grow a blister in its early stages. Theoretical predictions of blister growth based on the PECs theory are in excellent agreement with experimental results [3, 4]. This hypothesis motivates the detailed and advanced development of PECs-based theories contained in this work to understand the blister mechanics in thin films in general with particular attention to TBCs. First, the PECs-based theories are developed for the cooling rate-dependent spallation behaviour of alumina scales grown by oxidation on FeCrAl substrates. Consideration is given to the non-uniformity of plastic relaxation, cooling rate dependency, pockets of tensile stresses and spallation conditions. Then by using some experimental measurements of the height and radius of circular blisters, reported in the literature, the compressive residual stresses in the film and the fracture toughness of the film/substrate interface are determined by reversing the developed PECs-based theories. This work also develops another different technique together with a mechanical model to measure these same quantities, based on the blister morphologies in the circular blister test. The method works by considering the large mode mixity difference between the two cases of linear bending with small deflection and membrane stretching with large deflection. Second, monolayer telephone-cord blisters (TCBs) are considered: TCBs are blisters with wavy boundaries that propagate forward by the tip between the film and the substrate. By treating them as an assemblage of narrow straight-edged slices with a half-circular tip, PECs-based theories are developed in conjunction with the perturbation method to derive so-called 'Ω-formulae'. These formulae predict the four morphology parameters, namely, the local width and height, and the global wavelength and transverse amplitude. They all depend on the parameter denoted by Ω which represents the ratio between the plane-strain residual strain energy density and the interface fracture toughness; therefore, the name, 'Ω-formulae'. The quantity Ω is of high significance for the nucleation and development of TCBs. Next, to determine the compressive residual stress and interface fracture toughness, mechanical models are developed by using measurements of TCB morphology parameters published in the literature and reversing the Ω-formulae. Note that the PECs-based theories developed for TCBs also provide some physical understanding of 'branched' TCBs and 'web blisters'. Third, the PECs-based theories are extended to apply to multilayer coating/substrate material systems by considering through-thickness variable Young's modulus, Poisson's ratio and coefficient of thermal expansion. Mechanical models are developed for circular blisters, straight blisters and TCBs in multilayer or inhomogeneous films. These models provide insights to optimise the design of TBC material systems. Fourth, spallation tests are conducted using two types of TBCs on turbine blades to investigate their blister mechanics: One with a Pt-modified aluminide bond coat, and another with a Pt-diffused bond coat. Three-dimensional digital image correlation and several material characterisation techniques are used to examine the process of blister growth and spallation and to investigate the evolution of materials and microstructures after thermal ageing. Details of the nucleation, growth and development of blisters on the convex surfaces of turbine blades are presented for the first time. Furthermore, the percolation and coalescence of PECs are correlated with the microstructure of the bond coat close to the interface. The PECs theories and the PECs-based mechanical models developed in this thesis have been thoroughly validated against either independently obtained experimental results or results obtained by the author from the spallation tests of TBCs on turbine blades. Excellent agreement is observed, which provides strong support for the PECs hypothesis. It is therefore concluded that the PECs hypothesis and the developed PECs-based theories do provide a framework and the understanding to make valuable improvements in the design and manufacture of the monolayer and multilayer film/substrate material systems, including TBCs.