The Batak Toba character is one of the characters in the Batak Letter family (Batak character) which is used by the Batak Toba ethnic community to write the Batak Toba Language. Similar to the character that are related to it, the characters and spelling are not too different from other character such as Karo, Dairi, Mandailing or Simalungun. The use of the Batak Toba character is still very limited, also the introduction to the character. This study aims to recognize the Batak Toba character and to determine the accuracy of the k-nearest neighbor algorithm by implementing it to recognize the Batak Toba character. In the k-nearest neighbor algorithm there are two stages of implementation namely the training process and testing process. These two processes have different stages that need to be executed to obtain the right classification results and in accordance with the character of the Batak Toba character. By using KNN, it was found that the accuracy for each Batak Toba character that was successfully identified by using a collective method was around 72-90%