Describing and then explaining individual behavior during migration can help us to understand why (on both proximate and ultimate levels) birds migrate; the altitude(s) at which migratory birds fly, for example, can have far-reaching consequences. However, to date, no fine-scale, full-flight altitude data have been available for small (<100 g) migratory birds. We tracked 7 Swainson's Thrushes (Catharus ustulatus) carrying altimeters from takeoff until landing or near-landing during 9 migratory flights. The average recorded flight altitude for the 9 flights was 673.0 ± 523.2 m (mean ± SD); average maximum flight altitude for the 9 flights was 1,199.5 ± 862.7 m (range: 319.2–2,744.5 m). Initial ascent rates (0.42 ± 0.15 m s⁻¹, n = 8) matched predictions; final descent rates were 0.55 ± 0.30 m s⁻¹ (n = 5). Contrary to expectations, the thrushes made numerous (9.33 ± 4.42), significant (>100 m) altitude adjustments during their flights (1.44 hr⁻¹), not including initial ascent and final descent. The repeated changes in flight altitude that we observed should cause these birds to use more energy than they would if they flew at or near a single altitude for several hours at a time. We speculate that these altitude modifications may result from variation in atmospheric conditions or from the birds descending toward anthropogenic light sources during the flights.