The ongoing SARS-CoV-2 pandemic has necessitated a dramatic increase in our ability to conduct molecular diagnostic tests, as accurate detection of the virus is critical in preventing its spread. However, SARS-CoV-2 variants continue to emerge, with each new variant potentially affecting widely-used nucleic acid amplification diagnostic tests. RT-LAMP has been adopted as a quick, inexpensive diagnostic alternative to RT-qPCR, but as a newer method, has not been studied as thoroughly. Here we interrogate the effect of SARS-CoV-2 sequence mutations on RT-LAMP amplification, creating 523 single point mutation "variants" covering every position of the LAMP primers in 3 SARS-CoV-2 assays and analyzing their effects with over 4,500 RT-LAMP reactions. Remarkably, we observed only minimal effects on amplification speed and no effect on detection sensitivity at positions equivalent to those that significantly impact RT-qPCR assays. We also created primer sets targeting a specific short deletion and observed that LAMP is able to amplify even with a primer containing multiple consecutive mismatched bases, albeit with reduced speed and sensitivity. This highlights RT-LAMP as a robust technique for viral RNA detection that can tolerate most mutations in the primer regions. Additionally, where variant discrimination is desired, we describe the use of molecular beacons to sensitively distinguish and identify variant RNA sequences carrying short deletions. Together these data add to the growing body of knowledge on the utility of RT-LAMP and increase its potential to further our ability to conduct molecular diagnostic tests outside of the traditional clinical laboratory environment.