This systematic review by Adibaru Kifle Mulugeta, Durga Prasad Sharma, and Abebe Haile Mesfin aims to assess the application and usage of deep learning approaches in classifying and recognizing medicinal plant species. The review focuses on studies published between January 2018 and December 2022, identifying 31 studies that met the inclusion criteria. Key findings include: 1. **Geographical Distribution**: The studies were conducted in 16 different countries, with India leading in contributions (29%), followed by Indonesia and Sri Lanka. 2. **Dataset Preparation**: 67.7% of the studies used private datasets, often augmented and preprocessed. 3. **Feature Extraction**: 96.7% of the studies used plant leaf organs, with 74% focusing on leaf shapes. 4. **Deep Learning Methods**: 83.8% of the studies employed transfer learning with pre-trained models, and 64.5% used Convolutional Neural Networks (CNNs). 5. **Challenges and Opportunities**: The lack of globally available public datasets and the need for trustworthiness in deep learning approaches for medicinal plant classification are significant research gaps. The review highlights the importance of interdisciplinary collaboration and the need for further research, particularly in under-resourced countries, to address these challenges and advance the field of deep learning in medicinal plant species classification and recognition.This systematic review by Adibaru Kifle Mulugeta, Durga Prasad Sharma, and Abebe Haile Mesfin aims to assess the application and usage of deep learning approaches in classifying and recognizing medicinal plant species. The review focuses on studies published between January 2018 and December 2022, identifying 31 studies that met the inclusion criteria. Key findings include: 1. **Geographical Distribution**: The studies were conducted in 16 different countries, with India leading in contributions (29%), followed by Indonesia and Sri Lanka. 2. **Dataset Preparation**: 67.7% of the studies used private datasets, often augmented and preprocessed. 3. **Feature Extraction**: 96.7% of the studies used plant leaf organs, with 74% focusing on leaf shapes. 4. **Deep Learning Methods**: 83.8% of the studies employed transfer learning with pre-trained models, and 64.5% used Convolutional Neural Networks (CNNs). 5. **Challenges and Opportunities**: The lack of globally available public datasets and the need for trustworthiness in deep learning approaches for medicinal plant classification are significant research gaps. The review highlights the importance of interdisciplinary collaboration and the need for further research, particularly in under-resourced countries, to address these challenges and advance the field of deep learning in medicinal plant species classification and recognition.