The purpose of this study is that to investigate how different rounding rule in Angoff standard-setting judgments affect cut-scores. Four variables were manipulated in this study and they are rounding rules (nearest whole number, nearest 0.05, and nearest two decimal places), judgments (each individual item and cluster of items), item pools (full pool [180 items], Pool 1 [107 items], Pool 2 [109 items] and Three blocks of Items [53 items]), and cut-scores (basic, proficient and advanced), resulted in total of 72 conditions. A simulation design based on a data from the National Assessment of Education Progress was used to evaluate the effects of different rounding rules in different conditions. The results show that the nearest 0.05 and nearest two decimal places rounding rules worked fairly well than the nearest whole number rounding rule in respect to the average absolute biases and percentage above cut-score.