An estimate of ground-water availability in the Hawi area of north Kohala, Hawaii, is needed to determine whether ground-water resources are adequate to meet future demand within the area and other areas to the south. For the Hawi area, estimated average annual recharge from infiltration of rainfall, fog drip, and irrigation is 37.5 million gallons per day from a daily water budget. Low and high annual recharge estimates for the Hawi area that incorporate estimated uncertainty are 19.9 and 55.4 million gallons per day, respectively. The recharge estimates from this study are lower than the recharge of 68.4 million gallons per day previously estimated from a monthly water budget. Three ground-water models, using the low, intermediate, and high recharge estimates (19.9, 37.5, and 55.4 million gallons per day, respectively), were developed for the Hawi area to simulate ground-water levels and discharges for the 1990?s. To assess potential ground-water availability, the numerical ground-water flow models were used to simulate the response of the freshwater-lens system to withdrawals at rates in excess of the average 1990?s withdrawal rates. Because of uncertainty in the recharge estimate, estimates of ground-water availability also are uncertain. Results from numerical simulations indicate that for appropriate well sites, depths, and withdrawal rates (1) for the low recharge estimate (19.9 million gallons per day) it may be possible to develop an additional 10 million gallons per day of fresh ground water from the Hawi area and maintain a freshwater-lens thickness of 160 feet near the withdrawal sites, (2) for the intermediate recharge estimate (37.5 million gallons per day) it may be possible to develop an additional 15 million gallons per day of fresh ground water from the Hawi area and maintain a freshwater-lens thickness of 190 feet near the withdrawal sites, and (3) for the high recharge estimate (55.4 million gallons per day) it may be possible to develop at least an additional 20 million gallons per day of fresh ground water from the Hawi area and maintain a freshwater-lens thickness of 200 feet near the withdrawal sites. Other well-field configurations than the ones considered potentially could be used to develop more fresh ground water than indicated by the scenarios tested in this study. Depth, spacing, and withdrawal rates of individual wells are important considerations in determining ground-water availability. The regional models developed for this study cannot predict whether local saltwater intrusion problems may occur at individual withdrawal sites. Results of this study underscore the importance of collecting new information to better constrain the recharge estimates.