Neutral and ion measurements of the lunar environment appear to be conflicting. Two major discrepancies are the amounts of C+ and O+ measured from ion mass spectrometers compared to the stringent upper limits for C and O inferred from LRO measurements [Cook et al., 2013]. An additional puzzling feature is the derivation from ARTEMIS data that an ion species of mass~16 amu comes from an exosphere having a thermal scale height [Halekas et al, 2013], an observation that is not consistent with hot atomic oxygen from impacts and sputtering. We used Monte Carlo models of the neutral exosphere to compute abundances of many expected volatiles in the lunar exosphere, and based on these neutral models we computed the associated ion fluxes. We fitted these models to ARTEMIS data using an optimization method that also extracts confidence limits. Results suggest that the continuous presence of moderate amounts of water, hydroxyl, methane, carbon monoxide and dioxide in the lunar atmosphere must be posited in order to reconcile the ion and neutral budget of the atmosphere.