We investigate peer ability effects on high-stakes test scores at ages 16 and 18, and on the probability of university attendance. To account for endogeneity in peer ability, we use the average ability of the primary school peers of one’s secondary school peers, excluding those from the same primary school of the individual, as an instrumental variable for average secondary school peer ability. Our results show that average peer quality has a small effect on an individual’s test scores, and a larger proportion of low-quality peers has a significantly detrimental effect on achievements of an average student. Furthermore, peer ability seems to have a stronger effect on students at the bottom of the grade distribution, especially at age 16.