The diverse effects of emotion on human memory is a topic of discussion that has brought about much interest within psychology (Cardwell, Clark and Meldrum, 2004) with a particular emphasis on how further understanding of its influence can aid our knowledge about the structure of underlying memory systems. In recognition memory research, dual-process models propose that judgements can be based on two independent retrieval processes: “familiarity”, based on the perception of a memory’s strength, and “recollection”, founded on its contextual information. Much of the literature suggests that emotional stimuli, compared with neutral stimuli, can boost performance for contextual recollection, but not familiarity (Doerksen and Shimamura, 2001). The present study contained a source retrieval task, where presented scenes were of positive, negative or neutral valence with either a yellow or red border colour, in order to assess the contribution of emotion’s effects on recollection. The probabilities of correctly recognizing an item and retrieving its respective border colour were noted using a counter-balanced design. It was predicated that negative scenes would be more accurately recollected than both positive and neutral scenes, whereas item recognition performance would not be affected by emotional valence. Contrary to the hypotheses, results were indicative of an effect of valence on item recognition performance for the scenes, with negative images being remembered more accurately than positive and neutral stimuli. However, recollection for the border colours of the scenes was not found to differ across valence. The findings are discussed in respect to the Yonelinas Model of Dual-Processing (Yonelinas, 1999) to show the potential effect of emotion on recollective ability despite the results, and further study implications are suggested in order to establish the potential influence of recollection on the scenes using the current source retrieval procedure by applying additional testing methods such as a remember/know task.