Well information and groundwater-level measurements for the Columbia Plateau Regional Aquifer System in Washington, Oregon, and Idaho, were compiled from data provided by the U.S. Geological Survey and seven other organizations. From the full set of about 60,000 wells and 450,000 water-level measurements a subset of 761 wells within the aquifers of the Columbia River Basalt Group (CRBG) then was used to develop a simple linear groundwater-level trend map for 1968–2009. The mean of the trends was a decline of 1.9 feet per year (ft/yr), with 72 percent of the water levels in wells declining. Rates of declines greater than 1.0 ft/yr were measured in 50 percent of wells, declines greater than 2.0 ft/yr in 38 percent of wells, declines greater than 4.0 ft/yr in 29 percent of wells, and declines greater than 8.0 ft/yr in 4 percent of wells. Water-level data were used to identify groups of wells with similar hydraulic heads and temporal trends to delineate areas of overall similar groundwater conditions. Discontinuities in hydraulic head between well groups were used to help infer the presence of barriers to groundwater flow such as changes in lithology or the occurrence of folds and faults. In areas without flow barriers, dissimilarities in response of well groups over time resulted from the formation of groundwater mounds caused by recharge from irrigation or regions of decline caused by pumping. The areas of focus for this analysis included the Umatilla area, Oregon, and the Palouse Slope/eastern Yakima Fold Belt in the Columbia Basin Ground Water Management Area (GWMA) consisting of Adams, Franklin, Grant, and Lincoln Counties, Washington. In the Umatilla area, water levels from 286 wells were used to identify multiple areas of high hydraulic gradient that indicate vertical and horizontal barriers to groundwater flow. These barriers divide the groundwater-flow system into several compartments with varying degrees of interconnection. Horizontal flow barriers commonly correspond to mapped geologic structure and result in horizontal hydraulic gradients that progressively become steeper from north to south corresponding to an increase in structural complexity that may be impeding recharge from the uplands into the heavily developed areas. Most CRBG aquifers in the Umatilla area are declining and since 1970, cumulative declines range from about 100 to 300 feet. Significant vertical hydraulic gradients are documented for relatively small areas near Umatilla, and since the 1970s, downward vertical gradients in these areas have been increasing as hydraulic heads in the deeper units have declined. The absence of vertical gradients over much of the area may be a consequence of flow through commingling wells that results in the equilibration of the heads between aquifers. On the Palouse Slope in the central GWMA, large groundwater declines occurred during 1968–2009 along a north-south swath in the middle of the region. An analysis of 1,195 wells along major flow paths and through the area of persistent groundwater-level declines indicates that barriers to flow are not as evident in this area as in Umatilla. This is consistent with the geologic interpretation of the Palouse Slope as being a gently folded structure created by voluminous sheet flows of CRBG lavas. Groundwater discharge into the sediment-filled coulees, where the upper aquifers are intersected at land surface by incised canyons, is proposed as an alternative to explain local steepening of the hydraulic gradient along the Palouse Slope previously attributed to the presence of a groundwater dam. Comparison of generalized potentiometric surface maps developed for pre-development conditions and post-2000 conditions indicate that pre-development groundwater flow was from the uplands toward the Columbia and Snake River and that post-2000 flow patterns in the area are controlled by irrigation practices that have resulted in broad regions of elevated or depressed hydraulic head. In some cases, irrigation-related changes in head have reversed groundwater flow directions. Evidence of significant vertical hydraulic gradients exists, although much of the aquifer thickness is affected by commingling of wells. The effect of commingling and its relative contribution to problems related to groundwater-level declines remains unclear.