Older adults following an inpatient geriatric rehabilitation (GR) program commonly experience adverse health outcomes such as hospital readmission, institutionalization, and mortality. Although several studies have explored factors related to these outcomes, the influence of admission reason on the predictive factors of adverse health outcomes in the rehabilitation process remains unclear. Therefore, this study aimed to identify predictive factors for adverse health outcomes in inpatients attending GR according to their admission reason. This retrospective study included patients with orthogeriatric (OG) conditions and patients with hospital-associated deconditioning (HAD) admitted to GR after an acute hospitalization between 2016 and 2020. Patients were evaluated by a comprehensive geriatric assessment at admission, including sociodemographic data, social resources, clinical data, cognitive, functional and nutritional status, and physical performance measurements. Adverse health outcomes were collected (hospital readmission, institutionalization, and mortality). Univariate analyses and multivariate backward binary logistic regressions were used to determine predictive factors. In this study, 290 patients were admitted for OG conditions, and 122 patients were admitted due to HAD. In patients with OG conditions, lower Mini-Mental State Examination (MMSE) predicted institutionalization and mortality. Lower Mini Nutritional Assessment-Short Form predicted institutionalization, whereas lower Barthel Index and lower Tinetti-Performance-Oriented Mobility Assessment scores were associated with higher mortality. In patients with HAD, higher age-adjusted comorbidity index predicted hospital readmission and mortality, and lower Short Physical Performance Battery scores predicted institutionalization and mortality. Finally, lower MMSE scores, worse values in Older Americans Resources and Services Scale and male gender were associated with a higher risk of institutionalization. Predictive factors for hospital readmission, institutionalization, and mortality in patients with OG conditions and HAD during GR were different. Some of those predictors, such as nutritional status and physical performance, are modifiable. Understanding predictive factors for adverse outcomes, and how these factors differ by admission diagnosis, improves our ability to identify patients most at risk. Early identification of these patients could assist with prevention efforts and lead to a reduction of negative outcomes.