Reverse water-level responses in monitoring wells are widely known during pumping tests, where they are recognized as the Noordbergum or Rhade effects, and a similar response was observed in many of the 100+ interference slug tests conducted at a well field near Clemson, South Carolina. The reverse water-level effect is characterized by a drop in pressure head by 1 cm to several centimeters and it occurs in the first 10–100 s of the test. The reverse response is followed by a rise and fall of pressure head that is typical of slug-in tests. The reverse water-level response is highly repeatable, and it increases when the pressure used to create the slug test is increased. A conceptual model recognizes that opening displacement of the fracture wall can cause a pressure drop, even when the pressure increases in the wellbore. The conceptual model is supported by a closed-form, analytical solution and a numerical model that couple fluid pressure and deformation in a fracture. Characteristics of the reverse water-level response are sensitive to properties of the fracture system and enveloping formation. Parameter estimation methods can be used to invert theoretical analyses and use the reverse response to improve the characterization of fractured rock aquifers.