Human memory is a dynamic system which makes accessible certain memories of events based on a hierarchy of information, arguably driven by personal significance. Not all events are remembered, but those that are tend to be more psychologically relevant. In contrast, lifelogging is the process of automatically recording aspects of one’s life in digital form without loss of information. To be truly effective, lifelog software needs to be motivated by an understanding of cognitive psychology to inform the software feature set, so as to provide an effective search mechanism. This can be achieved by providing the user with multiple search axes such as date, time, location, people, common concept occurrence (e.g. driving, at computer, indoors, etc…) and organisation into a hierarchy of events. Over the last 4 years one of the authors has gathered a personal SenseCam archive of over 6 million images, including: all meetings, birthdays, social events, flights, meals, etc. over this time period. We have built search and browsing tools to allow effective user access to this collection, which are inspired by cognitive science. We contrast the theory behind these tools with our actual experiences in attempting to find select events of interest from a subset of 30,000 events from our collection. We show that even when presented with soundly motivated software, searching for events in a lifelog can still be a frustrating experience at times. We summarise the aspects of lifelog software that work well and those that require future development to avoid frustrating users.