Background: In software engineering there are several principles that got an impact on a software project. If these principles are applied the wrong way, or not considered, it can starve a software project. WIP-limit is of one those principles. WIP-limit is used to limit number of tasks people can work on. As of today, there is little evidence proving the impact of WIP-limit for software development. Aim: The aim of this work is to investigate the impact that WIP-limits have on software development. Methods: The methods used to investigate the research question were a case study of an in-house software development company. The case study was based on a data set with meta-data about each of the tasks that the software company worked on from 2008 to 2013. The data set was analyzed using an application developed for, and later described in this work. From the data set, the application measured variables such as WIP, throughput, bugs, lead time and churn for each team. The data produced by the application was interpreted with correlations and case summaries in statistical application. Correlation is a statistical method that measures how two variables change in relation to each other. Case summaries is a statistical method for grouping variables and calculate descriptive statistics. The correlation between variables is used to investigate the impact of WIP-limits. Results: Some of the results of this work were a mean correlation of 0.4 between WIP and throughput, a mean correlation of 0.2 between WIP and both bugs and lead time and a mean correlation of -0.1 between WIP and churn across the teams. Conclusion: Based on the data presented in this work, the conclusion is that WIP-limits matter in software development.