Mud gas logging has been part of mud logging for over 70 years and experienced various advances and improvements with ongoing technical evolution of gas extraction and measuring devices. Nevertheless interpretation of gas data is often ambiguous as an extensive array of drilling operations has the ability to influence the gas signal. Aim of the present study is to evaluate the possibility to utilize mud gas data from boreholes drilled by RAG in the Vienna Basin and the Molasse Basin as a tool to recognize hydrocarbon-bearing zones. In a first step, a normalization procedure applied by RAG to consider the influence of various drilling parameters (penetration rate, flow rate, bit diameter) on gas contents, has been critically evaluated. It turned out that differences in penetration rate have the greatest influence on normalized mud gas data. Partly, the normalization processes resulted in the elimination of gas shows related to gas reservoirs visible in raw data, partly it resulted in unrealistic high gas contents. This shows that absolute gas values should be interpreted only together with gas ratios. The influence of bit type, bit make up and bit diameter on the gas signal has been evaluated separately. The influence of these parameters is minor and partly compensated by lithostratigraphic influences. Because (raw and normalized) gas contents are often unreliable, the mud gas composition (C1/C2/C3/C(4+5) in relation to ΣC) has been calculated in addition. In contrast to mud gas logs showing gas contents, the gas composition logs are more sensitive to fluid content than to lithology. Gas composition data from nine wells (including sidetracks) drilled within and outside an oil field in the Molasse Basin suggest a relation between the fluid content in the Eocene reservoir and the gas composition logs of C2/ΣC versus C3/ΣC measured in the Lower Oligocene overburden. C3/C2 ratios above 1 are often observed above oil-bearing reservoirs. Uncertainties remain, mainly for C3+ components, related to fluctuation of the extraction coefficients of the utilized gas trap and comparable sampling precision. In the same oil field absolute gas values showed an increase at the boundary between the Zupfing and Eggerding formations. In the vicinity of the Schöneck Formation an increase of C1 compounds could be observed. Crossplots of mud gas data and petrophysical paramters (porosity, shale content, water saturation) were created. These show no correlation between gas contents and petrophysical parameters for different lithostratigraphic units or for rocks with different fluid contents. Moreover, mud gas data shows no major difference between "tight" and porous horizons. Comparing the mud gas composition of the Vienna Basin to that of the Molasse Basin shows a significant difference. In the Vienna Basin heavy hydrocarbon gases (C3+) are generally found in lower amounts than light hydrocarbon gases or are missing completely. In wells of the Molasse Basin C3+ compounds were continuously encountered in the deeper lithostratigraphic units. Some wells situated within the Imbricated Molasse showed greater amounts of C3 than C2 compounds in the mud gas, starting in the Zupfing Formation. The greatest potential to increase the quality of future mud gas interpretations, is seen in the correct gas trap position. Remaining questions about influences of operational parameters on mud gas could possibly be clarified directly at the well site, where direct communication with rig personal is an option. Correlating connection gas with connection depths could be used to verify the accuracy of the depth correlation of the gas signal. Connection gas itself could be useful as an indicator for overpressured, permeable horizons.