Clinical application of digital subtraction radiography is limited by the reproducibility in the orientation of the x-ray source, image receptor, and object. In this study, eight dental intraoral film alignment instruments (including five replicates of each) were tested for accuracy in repositioning over a period of 6 months. Each instrument was made by adhering one of six impression materials (including acrylics, compounds, and elastics) onto the bite blocks of commercially available alignment instruments. The dimensional accuracy and reproducibility of the orientation of the x-ray source with respect to the object were determined over time for each instrument by measuring the horizontal and vertical angulation change in the position of a buccally placed marker on a dried mandible with a lingually placed film grid. Nine measurements were made during the 6-month test period. The most reproducible instrument was the combination of Regisil, an elastic impression material, and a Rinn XCP bite block. This combination yielded a mean absolute horizontal angulation error of 1.34 degrees +/- 0.63 and a mean absolute vertical angulation error of 2.04 degrees +/- 0.82, yielding a total angulation error of 2.44 degrees +/- 1.16. This was within the acceptable range of accuracy needed to produce diagnostically useful information when digital subtraction radiography is used.