The results of a pilot randomised controlled trial of computerised cognitive training in older adults with mild cognitive impairment (MCI) are reported. Participants (N = 25) were randomised into either the treatment or waitlist training groups. Sixteen participants completed the 30-session computerised cognitive training program using exercises that target a range of cognitive functions including attention, processing speed, visual memory and executive functions. It was hypothesised that participants would improve with practice on the trained tasks, that the benefits of training would generalise to nontrained neuropsychological measures, and that training would result in improved perceptions of memory and memory functioning when compared with waitlist controls. Results indicated that participants were able to improve their performance across a range of tasks with training. There was some evidence of generalisation of training to a measure of visual sustained attention. There were no significant effects of training on self-reported everyday memory functioning or mood. The results are discussed along with suggestions for future research.