The task of collecting visitor data in an indoor environment and therein determining the occupancy of a building is an extensive task. Conventional methods are expensive, time-consuming, and often lack the ability to produce data in longer time series. Further, they often require disruption of the studied area as equipment must be deployed. The use-case for such data sets is often also limited as it can only reflect a certain state-of-time in the studied space.The thesis seeks to investigate if using a Wi-Fi tracking system as a methodology to measure building occupancy through passive data collection is a viable method. Through continuous monitoring over an extended period, it seeks to do a trend analysis over a limited time. The collected data reveals insights into peak usage periods and commonly used areas. Although not used in this study, this methodology could leverage existing Wi-Fi infrastructure eliminating the need of installing additional equipment.In the case study, temporary wireless access points were deployed in the office which was studied. Data was then gathered after a month-long measurement period. This data was analyzed, and patterns were discovered showing higher occupation in the beginning of weeks and declining towards weekends. The focus of the study was to see if the technology would work in this context as it had previously not been tested in office environments.The results showed that there were some differences between data predicted by the Wi-Fi tracking system and that observed by the authors while conducting manual counts for validation during certain hours in the office. This may stem from faulty calibration of the model or settings affecting the signal strength required for the system to register devices as visitors. Mainly the conclusions are bound to the thesis subject and not placed in a larger context, however applications in public transport are suggested.The study displays the possibilities offered by using Wi-Fi tracking systems as a method to collect and analyze data in indoor environments. Further study of the subject would likely find a better model calibration able to predict more accurate results. Such results could be used and integrated with HVAC control systems to contribute to energy savings.