Mentoring programmes can improve the attainment and retention of minoritized students in higher education, but schemes vary widely in their design and delivery. Universities have been urged to measure the effect of such programmes and to identify their most effective features at a local level. PEMENTOS (PEer MEntoring TO Succeed) was designed as such. Mentors and mentees were asked to arrange a number of in-person, one-hour “core meetings”. Mentors attended a mid-point review meeting, and both mentors and mentees were invited to attend an end-of-programme meeting. Pre- and post-intervention surveys showed that the sampled mentees were statistically significantly more confident and less worried about university life at the end of the programme than they were to start with. The biggest increases in confidence were seen in response to the statements “I feel confident about managing my time” and “I feel confident about planning my revision”, and the biggest reductions in worry were seen in response to the statement “I feel worried about adapting to university life” and “I feel worried about how to meet people and make friends”. Non-white mentees reported a greater increase in confidence about socializing with their peers and about managing their finances than white mentees did. We found that mentees logged in to our VLE at almost twice the rate of non-participants, and the fraction of assessment submissions was higher for mentees than for mentors and non-participants. Class attendance was also higher for mentees than non-participants, painting an encouraging picture of enhanced engagement among PEMENTOS participants.