To overcome the challenges faced by people with neurodegenerative diseases, Brain-Computer Interface (BCI) systems must make use of datasets relevant to patient's spoken languages. However, BCI research frequently faces setbacks due to the absence of such datasets for the target language population. This paper deals with the BCI datasets for one of the major Indian languages, Telugu, used by more than 90 million people, yet to obtain essential BCI datasets capturing the linguistic characteristics. To solve the unavailability of the Telugu BCI datasets, we created a dataset featuring EEG signal samples corresponding to frequently used Telugu words, aiming to fill this void and facilitate advancements in BCI research for native Telugu speakers. Through the utilization of Machine Learning and other classifiers, BCI systems can potentially translate and classify EEG signals into a display of or pronouncing Telugu words. Our dataset consists of both vocal and sub-vocal datasets of Telugu words and an English dataset of the corresponding English equivalents of the Telugu words. This IIST-BCI Dataset for Telugu is the first of its kind. It is dedicated to improving the accessibility of BCI technologies for Telugu-speaking individuals and fostering further research progress in this particular area.