<p>Twitter (now known as “X”) is a popular medium for Covid-19 related discussions. This paper presents a novel case study on sentiment analysis and topic modeling of Covid-19 vaccine-related tweets of users geo-located in India in the duration of 12 December 2020 to 11 November 2021 in the course of which more than half of the country’s 1.3 billion population got vaccinated. The sentiment analysis was performed, on day-wise basis, using unsupervised lexicon-driven sentiment analysis tools AFINN and Valence Aware Dictionary and sEntiment Reasoner, as well as BERTweet and Covid-Twitter-BERT transformer models pre-trained on Covid-19 tweets. The models were comparatively evaluated on a smaller annotated dataset of Covid-19 vaccine-related tweets prior to the longitudinal analysis. AFINN was ultimately chosen due to its better performance and ease of use for large unannotated data. AFINN analysis revealed that 51.38% of the 44,130 tweets were neutral, while 38.84% were positive, and 9.78% negative. Latent Dirichlet Allocation was used for topic modeling at the peak points corresponding to large sentiment fluctuations in the positive and negative longitudinal graphs derived using AFINN. The positive and negative vocabularies at peak points were scrutinized to derive insights on national/international events that triggered a change in public opinion. These findings could guide policy makers in gathering intelligence on misinformation and associated sentiments, and planning counter-measures to combat anti-vaccine campaigns. This study informs future strategies to counter vaccine hesitancy through targeted communication aimed at vulnerable groups that experience high anxiety and psychosocial burden during the pandemic.</p>