Assessment of the quality of translation is one of the important activities in translation. The translation assessment can be seen from three aspects: accuracy, clarity, and naturalness. This study aims to reveal the accuracy and fatigue of the translation results of students' Arabic-Indonesian official manuscripts, considering that researchers rarely touch the spiciness aspect than the translation's accuracy and reasonableness. This research uses the descriptive analysis method and content analysis model. The research sample is six graduate students in Arabic Language Education at the Indonesia University of Education. The data collection techniques used were observation, interviews, documentation, and questionnaires. From the results of this study, the researchers obtained several findings, namely: first, the level of mastery was very high, spread as many as eight data (44.4%), the moderate level of intelligence was spread as many as six data (33.3%), and the level of intelligence lacked as many as four data. (22.2%). Second, the level of fatigue in the translation of the official text for high levels of friction is as many as ten data (55.5%), there are five data (27.7%) for moderate fatigue, and three data (16.6%) is less tired.  So, from the results of this study, it is concluded that fatigue dominates as much as ten data (55.5%) compared to the very high level of mastery where there are only eight data (44.4%).