 This paper investigates the sentiment of tweets related to the COVID-19 pandemic in Saudi Arabia. It uses convolutional neural networks, CNNs, and bi-directional long-short-term memory, bi-LSTM, deep-learning algorithms to classify the sentiment of Arabic tweets. The results show that the performance of CNN was 92.8 percent, while the performance of bi-LSTM was 91.9 percent. Additionally, the study found that there was an increase in negative sentiments during the pandemic compared to the negative sentiments prior to the pandemic. This article was authored by Arwa Al-Kharni and Adaraman.