A two-dimensional, cross-sectional, finite-difference, ground-water flow model and a simple application of Darcy?s law were used to quantify ground-water flow (from a wetlands) beneath Levee 30 in Miami-Dade County, Florida. Geologic and geophysical data, vertical seepage data from the wetlands, canal discharge data, ground-water-level data, and surface-water-stage data collected during 1995 and 1996 were used as boundary conditions and calibration data for the ground-water flow model and as input for the analytical model. Vertical seepage data indicated that water from the wetlands infiltrated the subsurface, near Levee 30, at rates ranging from 0.033 to 0.266 foot per day when the gates at the control structures along Levee 30 canal were closed. During the same period, stage differences between the wetlands (Water Conservation Area 3B) and Levee 30 canal ranged from 0.11 to 1.27 feet. A layer of low-permeability limestone, located 7 to 10 feet below land surface, restricts vertical flow between the surface water in the wetlands and the ground water. Based on measured water-level data, ground-water flow appears to be generally horizontal, except in the direct vicinity of the canal. The increase in discharge rate along a 2-mile reach of the Levee 30 canal ranged from 9 to 30 cubic feet per second per mile and can be attributed primarily to ground-water inflow. Flow rates in Levee 30 canal were greatest when the gates at the control structures were open. The ground-water flow model data were compared with the measured ground-water heads and vertical seepage from the wetlands. Estimating the horizontal ground-water flow rate beneath Levee 30 was difficult owing to the uncertainty in the horizontal hydraulic conductivity of the main flow zone of the Biscayne aquifer. Measurements of ground-water flows into Levee 30 canal, a substantial component of the water budget, were also uncertain, which lessened the ability to validate the model results. Because of vertical flows near Levee 30 canal and a very low hydraulic gradient east of the canal, a simplified Darcian approach simulated with the ground-water flow model does not accurately estimate the horizontal ground-water flow rate. Horizontal ground-water flow rates simulated with the ground-water flow model (for a 60-foot-deep by 1-foot-wide section of the Biscayne aquifer) ranged from 150 to 450 cubic feet per day west of Levee 30 and from 15 to 170 cubic feet per day east of Levee 30 canal. Vertical seepage from the wetlands, within 500 feet of Levee 30, generally accounted for 10 to 15 percent of the total horizontal flow beneath the levee. Simulated horizontal ground-water flow was highest during the wet season and when the gates at the control structures were open.