Application of photonics is lagging behind IC because of two main drawbacks: size and manufacturing (particularly assembly) difficulties. MEMS technology allows to shrink the size of bulky optical components and to process optical components in batch with precise location, opening new opportunities for application. However, the traditional materials used for MEMS fabrication (Si and derived compounds) can not provide high quality optical components. I think that much better results could be obtained by using polymers which have already proved successfbl in terms of optical property and ease of fabrication. One such fabrication technique to make microlens using polymer is thermal reflow. The technique was first described by Popovic et a1 in 1988. It uses standard photolithographic process to create photoresist cylinders of desired dimension (based on the required focal length). These cylinders are then reflowed (melted) at an optimum temperature and time; such that the surface tension then pulls these cylinders into a hemispherical shape. The simplicity of the process and potential for large scale replication has attracted a lot of attention by various sectors of industry and research. However, a detailed study to precisely relate optical parameters (focal length, spherical aberration) with simple process parameters (dimension, temperature, reflow media, substrate, etc) is vital to get the desired profile of microlenses. I, thus studied various parameters governing the profile of microlenses for thermal reflow process. I also developed a mathematical model based on Young-Laplace equation to predict the profile formation, which aided to develop better insight about the process. In additon, I also used process modification to obtain other important categories of microoptical components,