A reflectometer probe was evaluated as part of a study designed to continuously monitor soil moisture and estimate ground-water recharge at an oil-spill site in a glacial outwash aquifer located near Bemidji, Minnesota. The two-prong, 30-cm-long probe is a self-contained "reflectometer" that does not require a time-domain reflectometry (TDR) cable tester to determine water content. That probe was compared to two three-prong TDR probes (30-cm-long and 50-cm-long) that require a cable tester. Campbell Scientific manufactures both probes (the two-prong probe is the CS615 and the three-prong probe is the CS605); the use of brand names herein is for identification purposes only and does not constitute endorsement by the U.S. Government. Three arrays of horizontally oriented probes were installed at one-half-meter depth intervals in the wall of a 2-meter-deep pit. Soil-moisture data were collected 6 times per day for 18 months in 1997-98 using a CR10X data logger. Results from field tests indicated that soil-moisture values measured with the CS615 probes were much more stable than those measured with the CS605 probes. For example, it is likely that variations in power supply voltage and line noise resulted in out-of-range (positive and negative) data using the CS605 probes. The data collected with the CS615 probes did not exhibit these anomalies. This loss of data critically affects recharge estimates, particularly during the spring. The data loss generally was intermittent, probe dependent, and represented less than 30 percent of all CS605 probe data collected. Recharge estimates based in part on the CS615 probe data (36 percent of 1997 precipitation) were similar to the 30-cm and 50-cm long CS605 probe data (43 and 35 percent of 1997 precipitation, respectively). Calibrations were run in the laboratory using the three types of probes installed in repacked columns of sandy sediments obtained from the field site. In the laboratory tests, volumetric moisture content measured using the factory-supplied calibration for the CS615 probes matched results based on gravimetric measurements to within about 0.02 cc/cc. Conversely, the moisture content versus dielectric-constant calibration curves measured using the 30-cm-long and 50-cm-long CS605 probes were shifted upward about 0.07 and 0.03 cc/cc, respectively, compared to Topp's standard calibration in the laboratory tests. Results indicate that the CS615 probe provides a dependable and accurate means for long-term monitoring of soil moisture in the glacial outwash being studied.