High-precision analyses of the isotopic composition of methane in ambient air can potentially be used to discriminate between different source categories. Due to the complexity of isotope ratio measurements, such analyses have generally been performed in the laboratory on air samples collected in the field. This poses a limitation on the temporal resolution at which the isotopic composition can be monitored with reasonable logistical effort. Here we present the performance of a dual isotope ratio mass spectrometric system (IRMS) and a quantum cascade laser absorption spectroscopy (QCLAS)-based technique for in situ analysis of the isotopic composition of methane under field conditions. Both systems were deployed at the Cabauw Experimental Site for Atmospheric Research (CESAR) in the Netherlands and performed in situ, high-frequency (approx. hourly) measurements for a period of more than 5 months. The IRMS and QCLAS instruments were in excellent agreement with a slight systematic offset of (+0.25±0.04)‰ for δ13C and (-4.3±0.4)‰ for δD. This was corrected for, yielding a combined dataset with more than 2500 measurements of both δ13C and δD. The high-precision and hightemporal- resolution dataset not only reveals the overwhelming contribution of isotopically depleted agricultural CH4 emissions from ruminants at the Cabauw site but also allows the identification of specific events with elevated contributions from more enriched sources such as natural gas and landfills. The final dataset was compared to model calculations using the global model TM5 and the mesoscale model FLEXPART-COSMO. The results of both models agree better with the measurements when the TNO-MACC emission inventory is used in the models than when the EDGAR inventory is used. This suggests that high-resolution isotope measurements have the potential to further constrain the methane budget when they are performed at multiple sites that are representative for the entire European domain.