We are often required to make decisions that can have safe or risky consequences. Evaluating the risk of each possible alternative is an important step before making our final decision. The main goal of the present research was to explore the neural basis of risk perception in a naturalistic context (driving). Twenty-two drivers evaluated the perceived risk in 72 traffic situations (previously categorized by driving instructors) while brain activity was recorded using fMRI. A neural network involving attentional factors, emotional processing, stimulus-response associations, and risk aversion was related to the perception of risks. Given the nature of our task, a more prominent role was played by emotional factors (evaluation of the consequences) than cognitive factors (e.g. probabilistic calculations). Moreover, activation in the insula, inferior frontal gyrus, precentral/postcentral gyrus, inferior parietal gyrus, and temporal and occipital regions linearly increased as a function of risk level. Our findings provide a new step towards understanding the neural processing underlying risk behavior in daily life tasks, which is particularly relevant given the study context and its important practical implications for our society.