We have examined processes affecting isotopes of argon (36Ar, 38Ar, 40Ar) in order to determine important atmospheric sources and sinks. Our simple model for argon evolution incorporates production of radiogenic argon in the mantle, outgassing of all argon species by extrusive and intrusive volcanism, and loss to space by knock-on sputtering above the exobase. Sputtering has been shown previously to be an important loss process for atmospheric species, especially isotopes of noble gases, which have few other mechanisms of escape. The integrated evolution of argon (36Ar, 38Ar, and 40Ar, respectively) is modeled in terms of these variables: (1) the planetary concentration of potassium, (2) the fraction of juvenile argon released catastrophically during the first 600 Myr., (3) potential variation in the time-history of sputtering loss from that suggested by Luhmann et al. <1992>, and (4) the volume of total outgassing to the surface as compared to outgassing contributed by volcanic release. Our results indicate that Mars has lost between 85--95% of 36Ar and 70--88% of outgassed 40Ar. Due to this substantial loss, the planet must have outgassed the equivalent of between 10 and 100 times the total volume of gases released by extrusive and intrusive volcanics. This indicates that volcanic outgassing, alone, is insufficient to explain the present-day abundances of 36Ar and 40Ar in the Martian atmosphere. Similar calculations for 20Ne suggest outgassed volumes of between 100 and 1800 times in excess of that due to volcanism. This results in a distinct Ne/Ar elemental fractionation, with a preference for outgassing argon, of the order of 10 to 17. Although the results must be evaluated within the model uncertainties, the results are compelling in that they unequivocally show the existence of additional sources of atmospheric volatiles and helps define a means to identify them. ¿ American Geophysical Union 1996