BACKGROUND: Standard nasopharyngeal swab testing for SARS-CoV-2 detection by PCR is not always feasible due to limitations in trained personnel, personal protective equipment, swabs, PCR reagents, and access to cold chain and biosafety hoods. METHODS: We piloted the collection of nasal mid-turbinate swabs amenable to self-testing, including both standard polyester flocked swabs as well as 3D printed plastic lattice swabs, placed into either viral transport media or an RNA stabilization agent. Quantitative SARS-CoV-2 viral detection by RT-qPCR was compared to that obtained by nasopharyngeal sampling as the reference standard. Pooling specimens in the lab versus pooling swabs at the point of collection was also evaluated. RESULTS: Among 275 participants, flocked nasal swabs identified 104/121 individuals who were PCR-positive for SARS-CoV-2 by nasopharyngeal sampling (sensitivity 87%, 95% CI 79-92%), mostly missing those with low viral load (<10^3 viral copies/uL). 3D-printed nasal swabs showed similar sensitivity. When nasal swabs were placed directly into an RNA stabilizer, the mean 1.4 log decrease in viral copies/uL compared to nasopharyngeal samples was reduced to <1 log, even when samples were left at room temperature for up to 7 days. Pooling sample specimens or swabs both successfully detected samples >102 viral copies/uL. CONCLUSIONS: Nasal swabs are likely adequate for clinical diagnosis of acute infections to help expand testing capacity in resource-constrained settings. When collected into an RNA preservative that also inactivates infectious virus, nasal swabs yielded quantitative viral loads approximating those obtained by nasopharyngeal sampling.