This study was to identify " the reality of using direct corrective feedback by teachers of Islamic subject at basic education schools-cycle 2" through given answers to certain questions. The study involved (36) male/female teachers of Islamic subject at basic education schools/ cycle 2 at Al-Batinah South, To answer the study questions, the researcher used two methods: 1. Interview. The interview included four points and each point led to certain questions. The interviews were recorded and then transcribed. After that a qualitative data analysis was done. 2. The use of an observation card contains corrective feedback types according to the translated model of Lyster & Ranta (1997). This method was used after videotaping samples of teachers lessons, given two lessons for each teacher (Recitation and another lesson of Islamic subject such as Interpretation, Hadeath, Jurisprudence (Figh) etc.) The major findings of this study were: 1. Among all types of corrective feedback, the most widely used type in classroom negotiation by teachers was Implicit Correction accounting for (47.86%) of the total frequently presented corrective feedback types. 2. The type that resulted in a high percentage of student uptake (Repair and Needs -Repair) was Elicitation which accounted for (97.78%) of uptake. 3. There were statistical differences at level (0.05) between the use of corrective feedback according to teacher's sex variable in only two types: Metalinguistic correction and repetition of student error, in favor of female teachers in both cases. 4. There were statistical differences at level (0.05) between the use of corrective feedback types according to lesson type variable among all types expect one type which was clarification request in favor of recitation lesson in all cases. The study made some recommendations as follows: There is a need to give variety to corrective feedback types rather than focusing mostly on one type as well as to encourage the use of corrective feedback types that promote students to self-repair and to partially repeat student's output to locate errors with added stress and intonation when using implicit correction. This type should beaccompanied by other corrective feedback types.