Spectroscopic observations of a large number of galaxies in the far-infrared are critical in the study of galaxy evolution. These observations are difficult at the moment, because of the lack of good far-infrared (FIR) detector technology. Microwave Kinetic Inductance Detectors (MKIDs) are showing promising results to fill this void. A common MKID design is the antenna-coupled coplanar waveguide (CPW) resonator patterned in a superconducting film. These resonators are extremely sensitive to changes in the Cooper pair density. FIR photons are energetic enough to break Cooper pairs. This means a superconducting resonator can be used as a pixel for a FIR camera. The inherent multiplexing advantage of MKIDs means they can easily be fabricated in arrays of up to 5000 pixels. By putting such an array in the SAFARI instrument on the cooled SPICA telescope a blind spectroscopic survey that can detect Milky Way sized galaxies out to redshift z = 2 will be possible. However, to achieve background limited performance on a cooled telescope like SPICA the sensivity of MKIDs needs to be improved by at least an order of magnitude. This project aims to investigate if this can be achieved by reducing the width of the CPW to much below what has typically been made using optical lithography (> 1 ?m). CPW resonators with a central line width as narrow as 300 nm were made in NbTiN using electron beam lithography and reactive ion etching. Using this fabrication method NbTiN resonators with a central line width varying between 0.3 ?m and 3 ?m were made. These resonators were used to do a systematic study of the width dependence of the kinetic inductance fraction, quality factor, responsivity, read-out power, noise and NEP. It is shown that the width dependence of all these properties does not change down to 300 nm. Encouraged by these results two prototype submicron pixel arrays were made in aluminum. These showed sensitivities comparable to resonators with a few micron wide CPW. The main reason for this lack of sensitivity improvement is the lower maximum read-out power of submicron wide aluminum resonators. Based on these results power handling seems a key issue that will determine if submicron resonators are a viable route to improve the MKIDs sensitivity. These results of this project indicate that aluminum has a different width dependence of its power handling compared to NbTiN. Therefore, a systematic study of submicron resonators in aluminum, which focuses on the power handling, is required.