Quantitative estimates of ground-water flow directions and traveltimes for advective flow were developed for the regional aquifer system of the eastern Snake River Plain, Idaho. The work included: (1) descriptions of compartments in the aquifer that function as intermediate and regional flow systems, (2) descriptions of pathlines for flow originating at or near the water table, and (3) quantitative estimates of traveltimes for advective transport originating at or near the water table. A particle-tracking postprocessing program was used to compute pathlines on the basis of output from an existing three-dimensional steady-state flow model. The flow model uses 1980 conditions to approximate average annual conditions for 1950-80. The advective transport model required additional information about the nature of flow across model boundaries, aquifer thickness, and porosity. Porosity of two types of basalt strata has been reported for more than 1,500 individual cores from test holes, wells, and outcrops near the south side of the Idaho National Engineering Laboratory. The central 80 percent of samples had porosities of 0.08 to 0.25, the central 50 percent of samples, O. 11 to 0.21. Calibration of the model involved choosing a value for porosity that yielded the best solution. Two radiologic contaminants, iodine-129 and tritium, both introduced to the flow system about 40 years ago, are relatively conservative tracers. Iodine- 129 was considered to be more useful because of a lower analytical detection limit, longer half-life, and longer flow path. The calibration value for porosity was 0.21. Most flow in the aquifer is contained within a regional-scale compartment and follows paths that discharge to the Snake River downstream from Milner Dam. Two intermediate-scale compartments exist along the southeast side of the aquifer and near Mud Lake.One intermediate-scale compartment along the southeast side of the aquifer discharges to the Snake River near American Fails Reservoir and covers an area of nearly 1,000 square miles. This compartment, which receives recharge from an area of intensive surface-water irrigation, is apparently fairly stable. The other intermediate-scale compartment near Mud Lake covers an area of 300 square miles. The stability and size of this compartment are uncertain, but are assumed to be in a state of change. Traveltimes for advective flow from the water table to discharge points in the regional compartment ranged from 12 to 350 years for 80 percent of the particles; in the intermediate-scale flow compartment near American Falls Reservoir, from 7 to 60 years for 80 percent of the particles; and in the intermediate-scale compartment near Mud Lake, from 25 to 100 years for 80 percent of the particles. Traveltimes are sensitive to porosity and assumptions regarding the importance of the strength of internal sinks, which represent ground-water pumpage. A decrease in porosity results in shorter traveltimes but not a uniform decrease in traveltime, because the porosity and thickness is different in each model layer. Most flow was horizontal and occurred in the top 500 feet of the aquifer. An important limitation of the model is the assumption of steady-state flow. The most recent trend in the flow system has been a decrease in recharge since 1987 because of an extended drought and changes in land use. A decrease in flow through the system will result in longer traveltimes than those predicted for a greater flow. Because the interpretation of the model was limited to flow on a larger scale, and did not consider individual wells or well fields, the interpretations were not seriously limited by the discretization of well discharge. The interpretations made from this model also were limited by the discretization of the major discharge areas. Near discharge areas, pathlines might not be representative at the resolution of the grid. Most improvement in the estimates of ground-waterflow directions and travelt