INTRODUCTION: Since their emergence, SARS-CoV-2 variants of concern (VOC) B.1.1.7 and B.1.351 have spread worldwide. We estimated the risk of hospitalisation and admission to an intensive care unit (ICU) for infections with B.1.1.7 and B.1.351 in Norway, compared to infections with non-VOC. MATERIALS AND METHODS: Using linked individual-level data from national registries, we conducted a cohort study on laboratory-confirmed cases of SARS-CoV-2 in Norway diagnosed between 28 December 2020 and 2 May 2021. Variants were identified based on whole genome sequencing, partial sequencing by Sanger sequencing or PCR screening for selected targets. The outcome was hospitalisation or ICU admission. We calculated adjusted risk ratios (aRR) with 95% confidence intervals (CIs) using multivariable binomial regression to examine the association between SARS-CoV-2 variants B.1.1.7 and B.1.351 with i) hospital admission and ii) ICU admission compared to non-VOC. RESULTS: We included 23,169 cases of B.1.1.7, 548 B.1.351 and 4,584 non-VOC. Overall, 1,017 cases were hospitalised (3.6%) and 206 admitted to ICU (0.7%). B.1.1.7 was associated with a 1.9-fold increased risk of hospitalisation (aRR 95%CI 1.6-2.3) and a 1.8-fold increased risk of ICU admission (aRR 95%CI 1.2-2.8) compared to non-VOC. Among hospitalised cases, no difference was found in the risk of ICU admission between B.1.1.7 and non-VOC. B.1.351 was associated with a 2.4-fold increased risk of hospitalisation (aRR 95%CI 1.7-3.3) and a 2.7-fold increased risk of ICU admission (aRR 95%CI 1.2-6.5) compared to non-VOC. DISCUSSION: Our findings add to the growing evidence of a higher risk of severe disease among persons infected with B.1.1.7 or B.1.351. This highlights the importance of prevention and control measures to reduce transmission of these VOC in society, particularly ongoing vaccination programmes, and preparedness plans for hospital surge capacity.