The decision about how and when to open schools to in-person instruction has been a key question for policymakers throughout the COVID-19 pandemic. The instructional modality of schools has implications not only for the health and safety of students and staff, but also student learning and the degree to which parents can engage in job activities. We consider the role of instructional modality (in-person, hybrid, or remote instruction) in disease spread among the wider community. Using two-way fixed effect (TWFE) specifications, we find evidence that instructional modality does lead to increases in COVID-19 spread in communities with moderate to high levels of preexisting COVID-19 cases, although the causal effect is small in magnitude. Recent research has indicated that these TWFE estimates could be biased under heterogenous treatment effects, however, we test for parallel pre-trends that suggest our estimates are unlikely to suffer from this type of bias. We further estimate event study models that show, in Michigan, but not in Washington, there is an initial increase in COVID-19 rates in the months after districts begin offering in-person schooling after which effects slowly fade out. Finally, there is little consistent evidence of differential effects in districts with lower or higher percentages of students reported to be in school buildings.