The intelligent networking of machines plays a crucial role in Industry 4.0. As a result, many machines are equipped with condition monitoring systems. Sensor data recorded on the machine is used to make statements about the condition of the machine or individual critical components. Condition monitoring detects damage at an early stage, so that unexpected machine failures can be prevented. For machine engineers, this results in added value in terms of operational safety as well as a financial advantage. A gearbox malfunction can lead to an entire machine coming to a standstill. For this reason, gearboxes are particularly suitable for condition monitoring during operation. The main failure mechanisms of gearboxes show a gradual progression of damage. The focus of current research at the Institute for Machine Elements at the University of Stuttgart is on pitting damage. In cooperation with SONOTEC, a measurement technology manufacturer specializing in ultrasound, pitting damage on gears was investigated. In the process, damage could be detected well before the failure criterion standardized in DIN 3990-5 which is of an area fraction on a tooth flank surface of 4%. This is possible with suitable sensor technology. In a series of tests, synthetic pitting damage with area fractions between 0.4 % and 2 % of the area of a tooth flank was examined. In addition to the size of the pitting damage, the speed, torque, and lubricating oil viscosity were systematically varied in the tests. In this way, the detection capability of the pitting damage could be tested under different boundary conditions. Different accelerometers were used to collect data. Compared to the accelerometers, the pitting damage is clearly visible in data of the T20 structure-borne sound probe from SONOTEC, particularly in the ultrasonic frequency range above 16 kHz. Thus, it is possible to reliably detect even the smallest examined pitting damage with area fractions of 0.4% of the area of a tooth flank.