A wind-tunnel test section with closed upper wall, slotted side walls, and open lower boundary was found theoretically to produce zero tunnel-boundary lift interference on a small wing with horizontal wake mounted at the center of the test section. For this test section the variation of the interference with angle of the vortex wake behind a high-lift-coefficient model was not large. Because of the small slot widths required for zero interference and of the effects of boundary layer, the theory is regarded as unreliable for predicting the slot widths required; however, the variation of the interference with the slot width for widths somewhat greater than those needed for zero interference was found to be small. The interference in the region likely to be occupied by the tall of a model was investigated in some detail and was found to change with slot width and with wake angle more strongly than did the interference at the lifting element. A limited investigation of the lift interference in a test section with closed upper wall and slotted side and lower boundaries was made to obtain the theoretically indicated slot width required for zero lift interference at a center-mounted wing with the wake horizontal.