Photonic crystals (PCs) are periodic structures that can prohibit the propagation of light with frequencies inside the photonic bandgap.[1] Recently, PCs constructed from metals have attracted considerable attention because of some unique features that conventional dielectric PCs cannot provide.[2]For instance, calculations indicate that there exists a large and complete bandgap in the optical regime for face-centered cubic (fcc) lattices constructed from metallic or metallodielectric (the discontinuous cermet topology) building blocks with lattice constants on the sub-500 nm scale.[3] Realization of such a three-dimensional (3D) structure, however, has posed a great challenge for current lithography techniques. [4] In comparison, the self-assembly approach [5] that involves spontaneous crystallization of colloids from a suspension appears to provide a much simpler and less-expensive route to PCs. Many studies have reported on self-assembled PCs made of dielectrics such as polystyrene and silica [6] because these materials can be readily prepared as monodisperse spherical colloids of suitable sizes and in relatively copious quantities. Recently, a number of semiconductors (e.g., CdS, ZnS, Se, CdSe, Ag 2 Se, and TiO 2 ) have been prepared as monodisperse spherical colloids and further crystallized into fcc lattices with stop bands located in the visible regime. [7] To our knowledge, there are very few reports on the fabrication of all-metallic colloidal crystals that operate at optical wavelengths. The main challenge seems to originate from the difficulty in preparing highly monodisperse metal spheres with size variations below 5 % and diameters controllable in the range of 0.1-1 lm.[8]For the synthesis of monodisperse spherical metal colloids, most previous studies have been limited to the size range below 100 nm.[9]To partially overcome the technical barrier associated with the direct synthesis of spherical metal colloids with diameters larger than 100 nm, Wiley and co-workers, as well as several other groups, have made progress by demonstrating a template-directed method for generating spherical colloids from a number of metals.[10] The use of a template, however, has a number of intrinsic shortcomings or limitations. For example, the quantity of colloids that can be produced in each synthesis run is limited by the available pores in the template. In addition, any defects in the template would lead to the formation of colloids with poorly defined structures and shapes. Furthermore, additional steps are often required in order to separate the metal colloids from the template. In another approach, coating of silica or polystyrene spherical colloids with metal shells has been explored with some success, [11] but the resultant core/shell particles often exhibit problems such as incomplete coating, uneven shell thickness, and deviation from a spherical shape due to surface roughness and anisotropic growth of the metal coating. We recently demonstrated two different, solution-based approaches (top-down and bottom-u...