We have, on behalf of the department of computer science at Karlstad University, designed and implemented a new version of an already existing system for anonymous Internet surfing. The original system, created by M. Reiter and A. Rubin at AT&T in the USA, is called Crowds. Since the existing system has proven impossible to take into practice one wanted to have a working system for research purposes. This report contains an account of the development of our version of the Crowds system and a detailed description of the existing system. We also give an introduction to the subjects of anonymity and cryptography. We have implemented the system in Java and have also made some tests regarding the usefulness of the system. We have found the system being useful for people who wish to be anonymous searching information on the Internet. Yet a person using the system must be prepared to do some compromises concerning usefulness and anonymity.