Plastic flow in bulk metallic glasses (BMGs) localizes into narrow bands, which, in the absence of a microstructure that could obstruct them, propagate unhindered under tensile loading. In constrained deformation conditions such as indentation and at notch roots, extensive shear band formation can occur. A key issue in the context of fracture of BMGs that is yet to be understood comprehensively is how their toughness is controlled by various state parameters. Towards this end, the change in fracture toughness and plasticity with short term annealing above and below the glass transition temperature, Tg, is studied in a Zr-based BMG. Elastic properties like shear modulus, Poisson's ratio as well as parameters defining the internal state like the fictive temperature, Tf, density, and free volume are measured and correlation with the toughness was attempted at. While the elastic properties may help in distinguishing between tough and brittle glasses, they fail to reveal the reasons behind the toughness variations. Spherical-tip nanoindentation and microindentation tests were employed to probe the size, distributions and activation energies of the microscopic plastic carriers with the former and shear band densities with the latter. Results indicate that specimens annealed at a higher temperature, Ta, exhibit profuse shear banding with negligible changes in the local yield strengths. Statistical analysis of the nanoindentation data by incorporating the nucleation rate theory and the results of the cooperative shear model (CSM), reveals that short term annealing doesn't alter the shear transformation zone (STZ) size much. However, density estimates indicate changes in the free volume content across specimens. A model combining STZ activation and free volume accumulation predicts a higher rate in the reduction of the cumulative STZ activation barrier in specimens with a higher initial free volume content. Of the macroscopic physical properties, the specimen density is revealed to be a useful qualitative measure of enhancement in fracture toughness and plasticity in BMGs. We turn our attention next to the brittle fracture in BMGs, with the specific objective of understanding the mechanisms of failure. For this purpose, mode I fracture experiments were conducted on embrittled BMG samples and the fracture surface features were analyzed in detail. Wallner lines, which result from the interaction between the propagating crack front and shear waves emanating from a secondary source, were observed on the fracture surface and geometric analysis of them indicates that the maximum crack velocity to be ~800 m/s, which corresponds to ~0.32 times the shear wave speed. Fractography reveals that the sharp crack nucleation at the notch tip occurs at the mid-section of the specimens with the observation of flat and half-penny shaped cracks. On this basis, we conclude that the crack initiation in brittle BMGs occurs through hydrostatic stress assisted cavity nucleation ahead of the notch tip. High magnification scanning electron and atomic force microscopies of the dynamic crack growth regions reveal highly organized, nanoscale periodic patterns with a spacing of ~79 nm. Juxtaposition of the crack velocity with this spacing suggests that that the crack takes ~10-10 s for peak-to-peak propagation. This, and the estimated adiabatic temperature rise ahead of the propagating crack tip that suggests local softening, are utilized to critically discuss possible causes for the nanocorrugation formation. The Taylor’s fluid meniscus instability is unequivocally ruled out. Then, two other possible mechanisms, viz. (a) crack tip blunting and resharpening through nanovoid nucleation and growth ahead of the crack tip and eventual coalescence, and (b) dynamic oscillation of the crack in a thin slab of softened zone ahead of the crack-tip, are critically discussed. One way of alleviating the fracture-related issues in BMGs is to impart a microstructure to it, which would either impede the growth of shear bands or promote the multiplication of them. One such approach is through the BMG composites (BMGCs) route, wherein a crystalline second phase incorporated in the BMG matrix. There is a need to study the effects of reinforcement content, size and distribution on the mechanical behavior of the BMGC so as to achieve an optimum combination of strength and ductility. For this purpose, an investigation into the microstructure and tensile properties of Zr/Ti-based BMG composites of the same composition, but produced by different routes, was conducted so as to identify “structure–property” connections in these materials. This was accomplished by employing four different processing methods—arc melting, suction casting, semi-solid forging and induction melting on a water-cooled copper boat—on composites with two different dendrite volume fractions, Vd. The change in processing parameters only affects microstructural length scales such as the interdendritic spacing, λ, and dendrite size, δ, whereas compositions of the matrix and dendrite are unaffected. Broadly, the composite’s properties are insensitive to the microstructural length scales when Vd is high (∼75%), whereas they become process dependent for relatively lower Vd (∼55%). Larger δ in arc-melted and forged specimens result in higher ductility (7–9%) and lower hardening rates, whereas smaller dendrites increase the hardening rate. A bimodal distribution of dendrites offers excellent ductility at a marginal cost of yield strength. Finer λ result in marked improvements in both ductility and yield strength, due to the confinement of shear band nucleation sites in smaller volumes of the glassy phase. Forging in the semi-solid state imparts such a microstructure.