A corona virus infectious disease (COVID – 19) is spreading quickly throughout the globe. An exhaustive data available for this infectious disease on confirmed cases, active cases, recovered cases and fatal cases and relation between them will give us some information for future statistical analysis and modelling. In this research paper, the focus is on Modelling and forecasting the COVID – 19 cases for India and Karnataka. With this view an auto regressive model has been developed for the confirmed and fatal cases which are validated for the period of data which has not been used for modelling. While comparing to the other existing models this gives preferable results.