Software testing is important to find errors in the software. So, software testing from the beginning of the development process is needed. To test software, identifying test cases is the main activity. To find errors in the software system, test cases are generated. This reduces software testing complexity and increases the software testing coverage. So, automating the software test case generation process reduces software development testing costs, time, personnel, and the risk of software project failures. Test case generation based on requirements is one of the ways to improve software quality. However, software test case generation is one of the most challenging and tedious tasks in SDLC. Most of the time, it is done manually, and traditional machine learning approach. However, newly emerged deep learning techniques with feature extraction improve the performance of software testing by identifying features and generating test cases automatically. As far as our knowledge is concerned, there is no research done on test case generation from Amharic use case description textual documents. Thus, we proposed a deep learning model to automate the test case generation process of software testing from the Amharic textual use case description document. Test case generation from the use case description of the Amharic text documents using the LSTM passes through the process of dataset preparation, preprocessing, word embedding, LSTM model building, training the model, and test case identification. We prepared and classified the dataset into 80% training set and 20% testing set and both sets passed through three text preprocessing steps (tokenization, normalization, stemming), and then feature extraction was done using word2vec. Extracted features are used as input for dep network building. To build our model we used LSTM deep learning techniques and we trained them. Our test case generator LSTM model generates test cases with 82.65% accuracy and 80.12 testing accuracy. Keywords: Deep learning, use case description, test case