COVID-19 has killed more than 500,000 people in the United States, as of mid-February 2021. Forecasting of COVID-19 spread is helpful for key policy discussions. The transmission coefficient kn of COVID-19 spread varies across time. Accurately forecasting COVID-19 spread is difficult because of the time-dependent kn and becomes more complicated when coronavirus vaccination needs to be considered. In this study, the l-i AIR model was further developed for analyzing COVID-19 spread accompanied by coronavirus vaccinations in the United States. We determined all values of kn prior to January 13, 2021 and calculated the actual number of cumulative infections (In) including asymptomatic infected individuals. We observed 4 plateaus of kn, which corresponded to four national social events. This suggests that events that reduce social distancing and/or percentage of mask wearing played an important role in the acceleration of COVID-19 transmission. Our simulations show that if the American people return to their normal life before 100 million of people are vaccinated, there is likely to be at least one large surge of daily COVID-19 cases. However, if the American people partially return (kn≤0.4) to normal life after 100 million vaccinations, and completely return (kn=1) to normal life after two thirds of the US population are vaccinated in addition to those who have gained some immunity through coronavirus infections, the US may avoid any additional major surge of COVID-19 cases.