This thesis analyzes the process and effect of database profiling and optimization in SQL Server. An example system was developed with a supporting database in which a large number of initial records and a large number of triggers were introduced. In this way, the corresponding load on the database management system was simulated, resulting in a respective increase in the response time to queries. By using tools for automatic profiling and optimization within SQL Server, corresponding performance improvements were achieved. In particular, recommendations based on the query execution plan and estimated costs were used to implement measures such as the creation of non- clustered indexes and statistics on the corresponding attributes in the database tables. A detailed analysis of the experimental results provided a better insight into the effect of automatic profiling and optimization. It is shown that the optimization suggestions are not successful in every scenario and that it is important to apply the suggested indexes and statistics judiciously. It was also found that the tuning advisor should be given a larger workload so that the generated recommendations have a global positive impact on the operation of the system.