Background <p>Thunderstorm asthma is an increase in severe asthma following thunderstorm events during high pollen conditions. However, sparse pollen measurements hinder epidemiological research of this phenomenon.</p> Objective <p>Is pollen measured at a single site predictive of thunderstorm asthma risk across a broad region?</p> Methods <p>We conducted a meta-analysis to estimate thunderstorm asthma risk on 19 city-level sites incorporating local weather and patient data but a single pollen site. We use meta-regression to explore effect modification by land cover and distance from pollen measurement location.</p> Results <p>Meta-analysis showed no evidence of a state-wide thunderstorm asthma effect. Meta-regressions suggest that increased vegetation was associated with higher thunderstorm asthma risk with reduced risk at greater distances from pollen collection sites.</p> Impact statement <p>The phenomenon of thunderstorm asthma in the U.S. remains poorly studied due to geographically sparse pollen collection sites. Using a 19-city study, we demonstrate that incorporating environmental characteristics, such as land cover of allergic-type pollen-producing grasslands and deciduous trees, can improve the prediction of thunderstorm asthma risk at far distances from pollen monitors. By increasing the precision of pollen estimates, we can improve the estimation of thunderstorm asthma human health risks and potentially optimize decisions for new pollen collection sites.</p>