This thesis aims to study if descriptive representation, i.e. the social likeness between representatives and population, is considered important and if it changes when a national parliament is prevented from convening in full-format during a crisis. A maintained descriptive representation is important as it reflects that representatives have equal opportunities to execute their representational duties during a crisis (fairness argument), and because the interests of underrepresented groups otherwise risk being represented to a lesser degree (interest argument). By adopting a mixed-methods approach, semi-structured interviews are used to study how representatives were selected to participate in the reduced voting procedures in the Swedish parliament during the COVID-19 pandemic, and if descriptive representation (gender, age, geographical) was considered in this process. Results show that while descriptive representation was considered by five out of eight party groups – most prominently gender representation motivated by the fairness argument – it was not as influential as other more practical factors in the selection process, such as proximity to parliament or the need of being present for other reasons. Furthermore, changes in the descriptive representation in the Swedish parliament are studied by voting data using descriptive statistics and regression analysis. Results show that the share of women increased slightly during votes during the pandemic, contrary to developments in many other parliaments. Representatives over 65 years old decreased, as did representatives from constituencies far from parliament. Significant differences in average voting participation rate arose for the different age and geographical groups, indicating unequal opportunities to execute their representational duties during the crisis.