This paper deals with photorealism in computer generated images. The question at issue and the cornerstone of my report has been - What techniques are available to achieve photorealism in computer generated images and how do you use these techniques? I have examined and tested a variety of techniques and methods used to render photorealistic images. The report is based on a number of test projects in which I ruled out different parts of a simplified production pipeline. In each project I have moved focus from the big picture to specific parts of the process and generated a short sequence and/or a number of pictures. Because of my choice to focus on rendering, I have used simple models and put less time into modelling and texturing. In my final results I think I’m pretty close to achieving photorealism, although several of the projects would have looked better and more convincing if more time were spent on them. After completing the thesis, I am convinced that rendering and compositing are the most important steps to render photorealistic images. I draw this conclusion based on my results where I have barely utilized modelling and texturing and still believe that my projects are on the way to photorealistic results.