The email communication system is threatened by unsolicited commercial email aka spam. In response, spam filters have been deployed widely to help reduce the amount of spam users have to cope with. This paper describes work towards helping users better understand the often complex decision making that is spam filtering. An investigation of a number of popular web-based email services suggests that the filtering process is typically implemented as a black box allowing very little user involvement. In order to explore how we could help users understand how spam filters work and how they assess messages we conducted a number of user experiments using a simulated email interface providing richer spam filtering information than the webmail interfaces we investigated. Feedback indicates that additional information provided by the interface would be welcome and suggests to further investigate ways to involve users in the filtering process.