In this thesis we describe a relatively new behavioral biometrics technique based on mouse interactions. Given software (by Behaviometrics AB) for capturing the user interactions from a pointing device, i.e. a mouse or a touchpad, a system has been developed which both abstracts and uses the behavioral data to verify the identity of a user. Using statistical pattern recognition techniques a series of tests are combined to distinguish between a genuine user and an imposter by comparing the results from the tests to a predefined accuracy level. The system is to be used in real time while a user is working at his computer as to be able to stop any intrusions while they are happening. In my experiments no concern has been taken to what the user is doing at his computer as to make the system viable for any and all applications. Experimental results show that the behavioral information obtained from a pointing device can be a great addition to the existing security software developed by the company Behaviometrics AB.