The neutral beam injectors of the tokamak experiment TEXTOR produce neutral particle beams in the megawatt range at 55 keV and up to a 10-s pulse length of the light atoms hydrogen, deuterium, 3He, and 4He for heating the fusion plasma of TEXTOR. The two injectors are equipped with one 5-MW ion source (plug-in neutral injector) each. The injected power of ~1.5 MW of each injector can be varied from 0 to 100% by opening the main beam target vertical aperture in steps of ~2 cm to the full opening of 50 cm. The symmetric truncation of the neutral beam profile at a target position 4.5 m from the ion source leads to no major deformation of the profile downstream at the entrance into the torus plasma at a 6-m distance from the ion source. Whereas usually the particle energy, i.e., acceleration voltage, and beam current or, alternatively, the gas pressure in the neutralizer at fixed energy must be varied to change the injected power, these beam parameters can be kept constant with the reported method to study the effect of different injected neutral beam powers on the fusion plasma. The transmitted power to the torus is detected by the calorimetrically measured remaining power on the beam target. The resulting transmitted neutral beam power as a function of the target aperture is in good agreement with the expected integral of the thus-truncated Gaussianlike beam profile, i.e., the error function. The scaling of the resulting injected neutral beam power, beam profiles, vertical full-width-at-half-maximum, and central power density with variation of the beam target aperture are in good agreement with the beamline simulation code PADET.