Low frequency noise (Flicker noise or 1/f noise) dominates the noise performance of most MOS amplifiers below a few kilohertz and often is the dominant noise source even at much higher frequencies. The understanding of low frequency noise, or 1/f noise, has been a difficult challenge and many areas of its origin and behavior are still not understood. As the device dimensions continue to shrink with each new generation of MOS technology, the effect of an individual defect on device performance becomes more pronounced. Bird's beaking is one of them. During the field oxide growth, encroachment into the active region effectively reduces the width of the transistor, which is called bird's beaking. For short channel devices, bird's beaking could be a significant part of the overall width of a transistor. In the past, most of the work on 1/f noise in MOSFETs published in the literature had focused on the effects of oxide traps near the interface, but no attention had been paid to the area of bird's beaking(side wall noise effects). We have evaluated 1/f noise performance of MOS transistor with and without bird's beaking. Typically, designers utilize the rectangular-shaped gate whenever possible since the rectangular geometry is convenient for layout, component density can be high and good models for this device have been developed. Nonrectangular devices (e.g. trapezoidal, toroidal, circular, 'V' shaped etc.) are occasionally used, however. We have evaluated flicker noise performance of nonrectangular transistors without bird's beaking (such as circular and concentric) and have determined that geometry (circular, concentric etc.) of the device does not affect the noise performance significantly for large device sizes. A special case study for 1/f noise in series connected MOS devices is also analyzed in this thesis. Finally, operating region and bias dependent flicker noise is discussed as well. Due to the importance of accurate noise measurement in analog and mixed signal applications, 1/f noise measurement techniques and guidelines are covered in great detail in this thesis. New algorithm, for accurate extraction of flicker noise co-efficient (K[Subscript f]) from experimental data, is also proposed and successfully demonstrated to show its effectiveness.