The present study aimed to reveal the practices of Arabic language teachers for non-native speakers in the treatment of written errors among learners. The study was applied to Arabic language institutes for teaching non-native speakers in Oman. The sample was 21. The study followed the descriptive method based on describing reality and its data. To the teachers of the Arabic language for non-Arabic speakers, after verifying its validity and consistency, the results proved that: The arithmetic averages of the paragraphs on the first component of the questionnaire (teacher practices for the preparation of the lesson) ranged between (3.63) a maximum, and (2.21) a minimum, as the study proved Pain The arithmetic averages for the second component (teacher practices during the implementation of the lesson) ranged between (4.86) a maximum and (1.34) a minimum, and the averages of the paragraphs for the third component (teacher practices during the evaluation of the lesson) ranged from a maximum of (4.84), And (3.41) as a minimum, the results also proved that there are no statistically significant differences between the mean responses of the study sample at the level of significance (,0.05) in the degree of their practice to treat the written errors in the learners due to the variables of study (type - experience - educational qualification). In the light of these findings, the study made a number of recommendations and proposals.