Minimum competency cut-off scores on a statistics exam were estimated under four conditions: the Angoff judging method with item data (n=20), and without data available (n=19); and the Modified Angoff method with (n=19), and without (n=19) item data available to judges. The Angoff method required free response percentage estimates (0-100) percent, whereas the Modified Angoff method used a seven point scale (5, 20, 40, 60, 75, 90, 95). Item data were in the form of frequency distributions of multiple-choice exam option selections for five student ability levels. Results indicated significantly: (1) greater intra-judge variability for the Angoff method; and (2) higher minimum competency cut-off scores for judges with performance data available for reference than for judges who did not have data available. Acceptable levels of inter-judge reliability, judge internal consistency, and item internal consistency were also observed. (Author)