Abstract Background COVID-19 patients seen in an emergency department (ED) are at high risk of complications including post-COVID-19 condition (PCC), commonly known as Long COVID. As evidence is emerging concerning the efficacy of early post-acute rehabilitation and therapeutic interventions, early ED identification supported by a clinical prediction rule, combined with appropriate outreach and health education, could contribute to alleviating the burden of the disease on health systems and positively impact the quality of life of those living with the post-COVID-19 condition. This study aimed to derive and validate a clinical prediction rule to identify adult ED patients at high risk of developing PCC three months after an acute infection. Methods and findings This derivation and validation study used data from an observational cohort recruited from 33 hospitals in five Canadian provinces participating in the Canadian COVID-19 Emergency Department Rapid Response Network (CCEDRRN). We included adults (age ≥18 years) with confirmed COVID-19 who presented to the ED of a participating site between October 18, 2020, and October 11, 2022. We randomly assigned participants to derivation (75%) or validation (25%) datasets, and prespecified clinical variables as candidate predictors. We used a fast step-down logistic regression to reduce the model to key predictors for our clinical prediction rule. Validation was planned only if the derived rule had an AUC of at least 80% to support clinically useful discrimination characteristics to separate those who will develop PCC from those who will not. Of 6,070 eligible patients, 2,511 (41.4%) reported PCC symptoms at three months. Our derived clinical prediction rule included nine risk factors (female sex, higher arrival respiratory rate, comorbidities (rheumatologic disorder and mental health condition), acute symptoms (sputum production, dizziness, diarrhea, chest pain, and fatigue)) and one protective factor (self-reported South Asian race). In derivation, the optimism-corrected area under the curve was 0.626 (95% confidence interval [CI] 0.610–0.643). Age and vaccination status were not retained in the final clinical prediction rule. The rule was only slightly better than chance and deemed not accurate enough to meaningfully guide decision-making in the ED. Therefore, we did not proceed to examine its performance in the validation cohort. Conclusions Despite rigorous methodology, we were unable to derive a clinical prediction rule with sufficient accuracy to predict PCC in emergency department patients at the time of the acute infection. However, we did identify several factors associated with the development of PCC that can guide future studies about the causes of PCC. The ambiguous nature of the current PCC diagnostic criteria and the extended follow-up pose challenges for deriving a useful clinical decision rule. Further research integrating comprehensive surveillance systems and biomarker data may also enhance prediction accuracy and refine personalized management strategies in the emergency department setting.