The paper attempts to forecast the growth pattern of the COVID-19 spread in India and examines the impact of the lockdown on its spread and deaths. Comparing different models for short-term forecasts—hybrid autoregressive integrated moving average with errorremodelling using fast Fourier transform—has been found to have better accuracy. It is observed that the data set starting from the first phase of the lockdown generates more accurate estimates. The impact analysis shows a clear trend break on 3 March for confirmed cases and 11 March for the deaths.
