We present a dataset of measurements and observations of combustible live and dead vegetation (also known as fuel structure) in eight stands impacted by hurricanes along the north shore of Prince Edward Island, Canada. We collected fuel structure data by applying the Next-Generation Fuels sampling protocol associated with the Next-Generation Canadian Forest Fire Danger Rating System. The main application for this database is to permit the comparison of fuel structure across the different stands. Additional applications include estimating potential fire behaviour under specific fire-weather conditions and informing local to national fuel type mapping when combined with remote sensing products.