Using data on Uruguayan adolescents, we estimate peer effects in risk attitudes. Relative risk aversion is elicited in an experimental setting. Identification is based on parents not being able to choose the class within the school of their choice. After controlling for school-grade fixed effect and addressing endogeneity due to simultaneity, we find a significant and quantitative large impact of peers on individuals risk aversion. An increase in one standard deviation of the group risk aversion produces an increase in 44-64% on an individual risk aversion. These findings enhance the importance of multiplicative effects related to risk behavior.