The performance of tile roofing assemblies as well as untreated cedar shake roofing assemblies exposed to continuous firebrand showers were compared. Specifically, experiments were conducted for two types of concrete tile roofing assemblies (flat and profiled), one type of terracotta tile roofing assembly (flat), and an untreated (without any fire retardant) cedar shake roofing assembly. The design of the roofing assemblies were based on construction guidelines in the USA. The duration of the firebrand flux was fixed at 20 minutes, and the wind speed was varied from 6 m/s to 9 m/s. These wind speeds were chosen to be able to compare roofing assembly performance to similar assemblies exposed to a batch-feed firebrand generator which had limited duration of firebrand exposure (6 min). The average firebrand mass flux that arrived at the surface of the roofing assemblies was 0.3 g/m2s Results indicated that for the untreated cedar shake assemblies, ignition occurred easily from the firebrand assault, and this type of roofing assembly generated their own firebrands after ignition. To attempt to quantify the degree of penetration, the number of firebrands that penetrated the tile roofing assemblies, and deposited onto the underlayment/counter-batten system was counted as function of wind speed for each assembly. Firebrand penetration was observed, even for the flat tile assemblies. It is believed that these are the first-ever experiments described in the peer-reviewed literature to expose wood and tile roofing experiments to continuous wind-driven firebrand showers.