This research compares the impacts of change requests due to requirement defects on the outcomes of software development projects developed using the Waterfall methodology. The three types of requirement defects examined are incorrect requirements, incomplete requirements and new requirements. Outcomes are measured in terms of total effort expended and software defects injected during software development. While prior literature has examined ways to minimize requirement defects, limited insights are available on the impacts of requirement defects that remain after baseline requirements have been gathered. A sample of 49 software projects following the Waterfall methodology from a large highly mature (CMMI level 5) software development organization was used to statistically estimate the hypothesized relationships between the variables. Using the coordination perspective to develop our model, we find that resolution of change requests due to new requirements increases defects injected as well as effort. The resolution of change requests due to incorrect requirements increases the number of new requirements as well as the number of defects injected. Resolution of change requests due to incomplete requirements do not have measurable impacts on software project outcomes. Efforts to minimize the number of change requests necessary due to new requirements, can therefore be an important factor in improving software project outcomes.