Multicollector noble gas mass spectrometry has yielded significant improvements in the precision of 40Ar/39Ar age determinations over the last decade [1,2]. We report initial results from the new Isotopx NGX mass spectometer at the WiscAr Laboratory that is equipped with 9 Faraday detectors and 1 ion counting multiplier. The NGX utilizes Isotopx’s ATONA® amplifier technology, enabling measurements spanning a dynamic range from below 10−16 A to above 10−9 A [3]. Moreover, the WiscAr NGX offers the unique opportunity to evaluate instrument performance using both a conventional, and a newer low temperature, Nier-type ion source, which allows for trap current variation. We have performed a series of tests to: (1) assess optimal measurement and integration times for blanks, baselines, and air aliquots of various sizes, (2) quantify the NGX sensitivity via measurements of first principle 40Ar/39Ar standards fused with a CO2 laser, and (3) compare the sensitivity and effects on instrument background between the conventional and new low temperature Nier-type sources. Initial results suggest that for most samples optimal precision is achieved when using 10 second integration times for measurements lasting 400-600 seconds. Measurements of low 36Ar signals are optimized with longer integration and measurement times. In addition to measurement optimization experiments, we performed a comparative analysis of both fusion and incremental heating data obtained using the WiscAr Nu Instruments Noblesse, MAP 215-50, and the NGX for a variety of geologic materials which range in age from Pleistocene to Permian. Pychron software [4] controls the entire NGX analytical system, including data collection and reduction. Careful experiment design in this automated system, when informed by optimized counting practices and knowledge of collector sensitivity, can lead to high precision 40Ar/39Ar dates. [1] Jicha et al. (2016), Chemical Geology 431, 54–66 [2] Mark et al. (2009), Geochemistry, Geophysics, Geosystems 10 [3] Cox et al. (2020), Geochron. Discuss, 2020-1. [4] Ross J., 2019, NMGRL/pychron v18.2, doi: 10.5281/zenodo.3237834.