Web-based software systems access some resources while executing the requests of the clients. Typically several requests arrive at the same time, thus, competitive situation is established for the resources. For modeling such situations queueing model-based approaches are widely recognized. There is a demand for researching the ways how performance models can become more efficient. This thesis addresses the issues to establish performance models and evaluation methodologies, which are suitable for performance prediction of web-based software systems. The first result group introduces and verifies queueing network models and evaluation algorithms to model multi-tier ASP.NET web-based software systems. This contribution also includes the identification and investigation of new dominant factors considering the response time and throughput performance metrics. The second and third contributions deal with the establishment and investigation of the evaluation and prediction techniques applying the identified thread pool and queue limit performance factors. These methods facilitate the efficient performance prediction of web-based software systems.