We explore possible links between North American Monsoon System (NAMS) seasonal (Jun-Jul-Aug-Sep) precipitation and pre-monsoon (previous autumn, winter, and spring) land surface conditions, including precipitation, temperature, soil moisture and snow cover anomalies. We hypothesize land and sea surface feedback mechanisms associated with NAMS precipitation, and we propose an approach for determining their dynamical links. Following previous investigators, we partition the NAMS region into four sub-regions (Monsoon West, South, North and East) based on the seasonality and variability of JJAS monsoon precipitation from 1961-1990, and evaluate the possible effects of previous land surface conditions in various subcontinental “predictor regions” on Monsoon West (MW), Monsoon South (MS) and Monsoon East (ME) monsoon precipitation. Data for the study were monthly aggregates from the extended retrospective Land Data Assimilation System (LDAS) archive for the period 1922 to 2001. The retrospective LDAS archive includes gridded precipitation (P), mean surface air temperature (Ts), and Variable Infiltration Capacity (VIC) land surface model-derived soil moisture (Sm), and snow water equivalent (SWE). Our preliminary results on MW indicate that land surface-monsoon relationships are not stable in time, as has suggested by past studies. For instance, we found a statistically significant negative relationship between winter (JFM) precipitation in the Southwest U.S. predictor region which includes southern California, Nevada, Utah, Arizona, western Colorado and New Mexico, and MW monsoon rainfall during the 1965-1990 period, but weak relationships for other periods. We also found negative correlations between MW precipitation and winter-spring SWE in a predictor region that included the mountainous portions of Utah and Nevada. These relationships were especially strong between 1965 and the late 1980s. Based on the concept that the onset of the NAM is dynamically induced by land-sea temperature contrasts, we hypothesize that for MW, more winter P leads to more winter and early spring SWE in the predictor area, hence more spring and early summer Sm, and lower spring and early summer Ts, which induces a weaker onset of the NAMS and vice versa. We find that for MW, the antecedent land surface link that we propose (SWE, Sm, and Ts) is stronger in the Utah and Nevada mountain source areas where SWE may play a significant role in underpinning the land surface memory effect into the atmosphere. We outline future work that will construct an exploratory seasonal monsoon precipitation predictive model based on antecedent conditions.