With the growing software industry and our dependency on complex software applications, it is vital that the software is free of faults. The objective of my study is to, identify and classify the most common type of defects that are committed during the course of a software development project in a real industrial setting. The defects data are collected from two live projects and are categorized into 4 different categories as per the nature of the faults. The data has then been analyzed to see which defect categories are committed most during the software development life cycle. The result from this study indicates that the category ‘Implementation Defects’ was significantly higher followed by ‘Incorrect/Extra functionality’, ‘Missing Information’ and ‘Ambiguous Information’. ‘Implementation defects’ were further categorized to analyze the main area that led to increased faults and established that ‘Semantic bugs’ were the common faults committed during implementation of an application.