Advances in digital technologies for information capture combined with massive in- creases in the capacity of digital storage media mean that it is now possible to capture and store much of one’s life experiences in a personal lifelog (PL). Information can be captured from a myriad of personal information devices including desktop comput- ers, mobile phones, digital cameras, video recorders, and various sensors, including GPS, Bluetooth, and biometric devices. The large personal archives that can be cap- tured using these devices create new opportunities such as the chance to gain more details on partially recalled life events, opportunities for self reflection, facilities to share experiences, the potential to find partially remembered facts, etc, but also pose new challenges to the research community, not the least of which is developing effec- tive means of retrieval. This thesis centers on the investigation of the challenges of retrieval in this emerging domain, and the proposal and evaluation of methods and algorithms which seek to meet these challenges. Methods to integrate implicitly recorded and derived context data types with content- based search in information retrieval (IR) algorithms for PL retrieval are developed. These algorithms focus on the use of an individual’s memories of items’ content and associated context data and on the use of implicit biometric indicators of items’ im- portance. These novel retrieval algorithms are evaluated over unique multimodal PL collections of 20 months duration. We find support for the use of recalled context data in retrieval using a novel algorithm which accounts for the structure of lifelog collections and user queries. We also find support for the use of individuals’ past bio- metric response associated with lifelog items to locate important items in lifelogs and to re-rank ranked retrieval result lists.