The state of a metallic glass is far from equilibrium. From an energy landscape (EL) perspective, [1] when a glass-forming melt is quenched into a disordered glassy state, it is trapped in one of many possible megabasins in the EL and inhabits one of a range of infinitesimally different structural configurations (inherent states), depending on its cooling rate and thermal history. When an external stress is applied, glasses with the same composition but different structural configurations respond differently, leading to differing mechanical behaviors being expected. Recently, both experimental observations and computational simulations [2][3][4][5][6][7][8][9][10][11][12][13][14] have shown that the plasticity of bulk metallic glasses (BMGs) is very sensitive to processing conditions, such as thermal history, casting temperature and post-casting heat treatment. Processing influences the plasticity as a result of changes in the microstructure, causing ordering over short distances, even though BMGs all show an apparently amorphous macroscopic behavior. [6,7] In practical applications, the values and reliability of mechanical properties are the most important factors; [8] however, there is little published work addressed the reliability of the mechanical properties for BMGs, [9,10] despite widespread reports of data scattering in mechanical properties tests, especially with regard to measurement of plasticity. An evaluation of the stability of plasticity values not only is important for practical applications but also would provide insight into the deformation mechanism.The Weibull distribution is commonly applied to crystalline alloys for evaluating the statistical stability of their mechanical properties, such as strength and fracture toughness. [8,11] A coefficient, called the Weibull modulus, m, is developed to indicate the variability of the material property: a low value of m indicating a high variability, and vice versa.In this paper, we examine the statistical characteristics of the mechanical properties, including yield strength and compressive plasticity, of a typical zirconium-based BMG. We show that the probability functions of both strength and plasticity follow Weibull distributions. The strength of the BMG is quite stable, comparable to crystalline metals and alloys, while its plasticity is much les stable. We find that the stability of the compressive plasticity is susceptible to the presence of free volumes, with the value being influence by variations in their fraction and distribution. These results are useful in understanding the large degree of data scattering observed in the plasticity of BMGs, as well as deformation mechanism.Figure 1(a) shows some typical results of the compression tests. The BMG specimens, even those with the same composition and similar casting conditions, exhibit a surprising wide range of plasticity values, varying from less than 5% to more than 30% plastic strain. For instance, some specimens show low plasticity of 5% (curve 1), while a few cases of significantly large ...