We disclose a novel class of nanoparticle, dubbed nanoporous gold nanoparticle (NPGN). As shown in Fig. 1, NPGN features a fine porous network with pore size ~ 20 nm throughout its entire volume, which is not seen in any existing gold nanoparticles including solid- or hollow-core nanosphere, nanorod, nanoshell, nanocrescent, and nanocage. The external shape of our first NPGN is similar to a nanodisk with diameter ~ 400-500 nm and 75 run thick. Both the diameter and thickness can be easily tuned by slightly changing fabrication parameters. The high porosity is intriguing and critically important in several aspects. First, the increased surface area would permit NPGN to carry a much higher payload of surface adsorbates. This feature has significant implication in nanoparticle-based molecular cargo for the delivery of drugs, proteins, DNA and RNA into cells. It has a significant potential in improving current cancer treatment via chemotherapy or radiation therapy or the combined. Second, the NPGN is "semitransparent" due to its porous nature, thus, the internal surface adsorbates would potentially be optically measured. In other words, the amount of internal payload can be quantified via optical methods. Third, with proper surface linkage, the entire 3-dimensional internal volume can be "filled" and thus further increase payload without paying the price of size increase. Fourth, due to the fine pore structures, the majority of the "filler" or surface adsorbate molecules are within the plasmonic field or "hot spots". We believe this is the fundamental mechanism giving rise to our recently observed intense Raman scattering from a benzenethiol self-assembled monolayer (SAM) coating. A heuristic argument similar to that in the discovery of surface-enhanced Raman scattering (SERS) is that the increase of surface area (~10-30X) cannot account for the ~ 4-5 orders of magnitude increase in SERS intensity by comparing solid-core gold nanodisk and NPGN. The porous nature must have modified the nanoplasmonic behavior dramatically. Another heuristic explanation can be applied to the red-shifted plasmon resonance peak. Colloidal gold nanosphere peaks at ~ 540 nm and is relatively insensitive to size. Red-shifted plasmonic peak is known to be present in solid-core gold nanodisk (peak ~ 700 nm) and un-pattemed, i.e., continuous, NPG thin film (peak~ 650 run) [1]. It appears that the combination of NPGN's shape and the fine porous network has further red-shifted the placmonic peak into the near-infrared regime, at least to 785 nm employed in our experiments. This further red-shift provides a strong indication of synergistic coupling between external shape of a nanoparticle and its internal nanostructures.