Every year an estimated 4–5 million migratory birds collide with communication towers in the United States. We examined the relative risks that tower support systems and tower height pose to migrating and other birds. We collected data comparing tower support systems (guyed vs. unguyed) and tower height categories in Michigan during 20 days of the peak of songbird migration at 6 towers in September–October 2003, 23 towers in May 2004, 24 towers in September 2004, and 6 towers in both May and September 2005. We systematically and simultaneously searched for bird carcasses under each tower and measured carcass removal and observer detection rates each season. Of those towers, 21 were between 116 and 146 m above ground level (AGL, medium) and 3 were >305 m AGL (tall). During the five 20-day sample periods we found a mean of 8.2 bird carcasses per guyed medium tower and a mean of 0.5 bird carcasses under unguyed medium towers. During four 20-day sample periods we detected a mean of 34.7 birds per guyed tall tower. Using both parametric and nonparametric tests (Mann—Whitney U—test, Kruskal—Wallis test, and Tukey's Honestly Significant Difference multiple comparison procedure) we determined that unguyed medium towers were involved in significantly fewer fatalities than guyed medium towers. We detected 54–86% fewer fatalities at guyed medium towers than at guyed tall towers. We found 16 times more fatalities at guyed medium towers than at unguyed medium towers. Tall, guyed towers were responsible for 70 times as many bird fatalities as the unguyed medium towers and nearly five times as many as guyed medium towers. These findings will provide managers and regulators, such as the US Fish and Wildlife Service, with quantitative data; thereby, allowing them to effectively work with the Federal Communications Commission in siting and authorizing tower placement.