The nanostructured surfaces, as seen in both nature and in the lab, offer a broad range of advanced functionalities, such as stunning structural colors, antireflective, self-cleaning, superhydrophobic, superhydrophilic or antifogging effects. Those effects are facilitated by the specific arrangement of micro- and nano- structures on the surface. Manmade nanostructured surfaces, formed by advanced microfabrication techniques, can often mimic or even exceed natural ones in some property. However, there is a substantial limitation, as most of the abovementioned techniques work only on a flat, planar surface. One of the most widely used fabrication techniques, polymer injection molding, is not only capable of replicating extremely small structures, but also to produce parts with complex, non-planar shapes, at affordable cost. If we want to use injection molding for making products with nanostructured surfaces, we need to fabricate molds facing the same problem. In this work, we address this problem. At first, we need to verify if the previously reported monolayer adhesion-reducing coating (FDTS) can be used under different conditions during actual injection molding. Such coatings are critical to facilitate de-molding of the nanopatterned parts. We analyzed the coated surfaces of aluminum, titanium, and nickel molds before and after 500 molding cycles, using X-ray photoelectron spectroscopy, AFM and contact angle measurements. We show that the contact angle and that the fluorine concentration on the surface remains increased. These results enable us to predict the coating lifetime and the linearity of the coating removal. Based on the data, we can state that FDTS can be used for coating of molds, and is particularly suited for coating of nanostructured molds. We can also rank tested metals in the order of expected lifetime in descending order: aluminum, titanium, and nickel. The second problem addressed is the forming of a nanostructured surface on a non-planar substrate. We used the hydrostatic nanoimprinting technique, with the HSQ films deposited by spin-coating and spray-coating methods, on flat and curved mold inserts. The HSQ films are durable and tough, which makes them good candidates for molds, but the room temperature nanoimprint such high viscosity films as the HSQ dictates the use of extreme pressures, up to 800 bar. We designed and tested a special device capable of operation at those pressures, and used it to transfer the precise nanopattern with a period of 426 nm onto HSQ films on spherical surfaces with radii of curvature as low as 500 μm. With the pattern transferred onto a curved substrate, we investigated the pattern distortion, resulting from contact between inherently flat nickel masters and double-curved spherical surfaces. The mean pattern period was measured as a function of radial distance and found to be in good agreement with the foil strain computed with a finite element (FE) method. Moreover, this FE method was able to predict the contact pressure as a function of the radial distance. The FE method was also able to predict the sudden pressure drop at the position corresponding to the experimentally observed limit behind which the pattern was no longer replicated. We demonstrated a feasible method to produce non-planar nanopatterned surfaces for use as injection mold inserts. This opens new possibilities for making affordable polymer products with functional nanostructured surfaces.