An evaluation method supported by robust statistical analysis was applied to historical measurements of Ar half-life. The method, based on the most frequent value (MFV) approach combined with bootstrap analysis, provides a more robust way to estimate Ar half-life, and results in MFV years with uncertainty corresponding to the 68% confidence level. The uncertainty is a factor of 3 smaller than that of the most precise re-calculated Ar half-life measurements by Stoenner et al. and a factor of 2.7 smaller than that of the adopted half-life value in nuclear data sheets. Recently, the specific activity of the beta decay of Ar in atmospheric argon was measured in several underground facilities. Applying the MFV method to a specific activity of Ar from underground measurements results in SA39Ar/Ar(MFV)=0.966-0.018+0.010Bq/kgatmAr with uncertainty corresponding to the 68% confidence level. In this paper the method to determine the half-life of Ar using the specific activity of Ar in atmospheric argon is also discussed.