Sweden was hit hard in the beginning of the COVID-19 pandemic with deaths per capita among the highest in Europe. The pandemic was a stressful time especially for healthcare workers caring for COVID-19 patients. Various studies have evaluated whether nurses caring for these patients had higher levels of acute stress, but typically with measures that either used older DSM-IV criteria for Acute Stress Disorder (ASD) or general measures of acute stress. We recruited an online sample (N = 101) of nurses in Sweden from COVID-19 specialized units (ICU), Emergency (ER), and other units (Other), and asked them to answer questionnaires retrospectively to the peak of infections in Sweden. We aimed to evaluate: 1) the psychometric properties of the translation of the Stanford Acute Stress Reaction Questionnaire-II (SASRQ-II, which follows DSM-5 criteria for ASD) into Swedish, 2) whether nurses in COVID-19 units had experienced more acute stress than nurses in other units, and 3) the extent of potential acute stress disorder. The SASRQ-II evidenced good construct, convergent and divergent validity, and good reliability. It showed that ICU nurses reported significantly more acute stress than the other two groups, a difference that could not be accounted for by demographic or other variables. A retrospective diagnosis of ASD using the SASRQ-II suggested that 60% of nurses might have fulfilled ASD criteria, but no differences across groups were found.