SUMMARYThe Apache Point Observatory Lunar Laser-ranging Operation (APOLLO) project began in 2006 using a 3.5 m telescope on a high peak in the Sacramento Mountains of New Mexico. It has been one of the best performing (in terms of returned photons per viewing period and range uncertainty) of a handful of similar sites worldwide. The purpose of installing a superconducting gravimeter (SG) in 2009 was to improve the determination of local deformation, directly affecting the telescope motions, as part of an effort to reduce the error in lunar distance to the mm level. We have now accumulated 10 yr of gravity measurements, together with data from a nearby permanent GPS Plate Boundary Observatory (P027) site. This paper describes the traditional analysis of SG data to produce a local tidal model and to determine a number of necessary dynamic corrections that are specific to the site, of which local atmospheric and hydrology attraction and loading are the most important. Loading corrections are available through the Ecole et Observatoire des Sciences de la Terre (EOST) website as part of the International Earth Rotation Service, and we consider 12 different EOST loading series for global atmosphere and hydrology loading and attraction. Nonetheless, local hydrology can only be addressed at the required level of detail using local models determined directly from the gravity data. We devote a major portion of the paper to the development of several water storage and conceptual tank models that succeed in reducing the large seasonal variance of the SG residuals by more than 90 per cent. Yet there remain episodes in the corrected residuals that depart by up to 4 $\mu {\rm{Gal}}\,$ from the hydrological models, and the mean error of our fit remains between 1.5–2.0 $\mu {\rm{Gal}}\,$ for our two data sets, 2009–2012 and 2013–2018 respectively. Unfortunately the vertical GPS data relies on the UNAVCO processing, and has typically poor vertical accuracy. We find the correlation between the GPS and gravity residuals to be correspondingly weak, and then discuss briefly the incorporation of the displacement and gravity data into the Planetary Ephemeris Program for solving for the lunar distance.