Car racing is a mentally and physically demanding sport. The track time available to train drivers and test car setups is limited. Race car simulators offer the possibility of safe, efficient, and standardized human-in-the-loop training and testing. We conducted a validation study of a race car simulator by correlating the fastest lap times of 13 drivers during training events in the simulator with their fastest lap times during real-world race events. The results showed that the overall correlation was .57 (p = .044). Next, the effect of brake pedal stiffness (soft: 5.8 N/mm vs. hard: 53.0 N/mm) on racing performance was investigated in the simulator. Brake pedal stiffness may have an important effect on drivers' lap times, but it is impractical to manipulate this variable on a race car during a real-world test session. Two independent experiments were conducted using different cars and tracks. In each experiment, participants (N = 6 in Experiment 1 and N = 9 in Experiment 2) drove alternately with the soft and hard pedal in eight 20-min sessions (Experiment 1) or six 15-min sessions (Experiment 2). Two hypotheses were tested: (1) the soft pedal yields faster cornering times for corners that include a long brake zone, and (2) the hard pedal yields more high-frequency brake forces. Experiments 1 and 2 confirmed the second hypothesis but not the first. Drivers were highly adaptable to brake pedal stiffness, and the stiff pedal elicited higher pedal forces and more high-frequent brake pedal inputs. It is concluded that the racing simulator is a valuable tool for driver assessment and for testing adoptations to the human–machine interface.