The implementation of isolation precautions for patients with suspected Coronavirus Disease 2019 (COVID-19) and pending test results is resource intensive. Due to the limited availability of single-bed rooms at our institution, we isolated patients with suspected COVID-19 together with patients without suspected COVID-19 on-site in multiple-bed rooms until SARS-CoV-2-test results were available. We evaluated the likelihood of SARS-CoV-2 transmission to individuals sharing the room with patients isolated on-site. This observational study was performed at the University Hospital Basel, Switzerland, from 03/20 − 11/20. Secondary attack rates were compared between patients hospitalized in multiple-bed rooms and exposed to individuals subjected to on-site isolation precautions (on-site isolation group), and patients exposed to individuals initially not identified as having COVID-19, and not placed under isolation precautions until the diagnosis was suspected (control group). Transmission events were confirmed by whole-genome sequencing. Among 1,218 patients with suspected COVID-19, 67 (5.5%) tested positive for SARS-CoV-2. Of these, 21 were isolated on-site potentially exposing 27 patients sharing the same room. Median contact time was 12 h (interquartile range 7–18 h). SARS-CoV-2 transmission was identified in none of the patients in the on-site isolation group vs. 10/63 (15.9%) in the control group (p = 0.03). Isolation on-site of suspected COVID-19-patients in multiple-bed rooms avoided single-room occupancy and subsequent in-hospital relocation for many patients without confirmed SARS-CoV-2-infection. The absence of secondary transmission among the exposed patients in the on-site isolation group allows for assessment of the risk/benefit ratio of this strategy given the limitation of a small sample size.