This meta-analysis was conducted to quantify the risk of patients exhibiting cognitive deficits in the acute phase of COVID-19 at the time of the first variants (i.e., before the vaccine) and quantify the potential vulnerability of older patients and those who experienced more severe respiratory symptoms. To this end, we searched the LitCovid and EMBASE platforms for articles, including preprints, and included all studies (n = 48) that featured a measurement of cognition, which encompassed 2233 cases of COVID-19. Of these, 28 studies reported scores on global cognitive efficiency scales administered in the acute phase of COVID-19 (up to 3 months after infection). We were able to perform a meta-analysis of proportions on 24 articles (Npatients = 943), and a logistic regression on 18 articles (Npatients = 518). The meta-analysis for proportion indicated that 52.31% of patients with COVID-19 exhibited cognitive deficits in the acute phase. This high percentage, however, has to be interpreted taking in consideration the fact that the majority of patients were hospitalized, and some presented neurological complications, such as encephalopathy. A bootstrap procedure with random resampling revealed that an age of 59 was the threshold at which one would be more prone to present cognitive deficits. However, the severity of respiratory symptoms did not influence the scores on a global cognitive efficiency scale. Overall, our results indicated that neuropsychological deficits were a major consequence of the acute phase of the first forms of COVID-19.