A long-standing paradigm in the ecology of the Alaskan taiga states that black spruce (Picea mariana [Mill.] BSP) replaces white spruce (Picea glauca [Moench] Voss) after several centuries of primary succession on floodplains. According to this Drury Hypothesis, autogenic thickening of organic horizons and shrinking of the active layer interact with the species' different physiological tolerances to cause black spruce dominance. We test the Drury Hypothesis on >200-year-old portions of the Tanana River floodplain near Fairbanks, Alaska, and reject it. In the meander belt portion of the study area, white spruce mixed with black spruce persists on geomorphic surfaces approximately 3 000 years old. Predictions of the Drury Hypothesis regarding active-layer and organic-horizon thicknesses are not substantiated. Neither of these variables correlates with the abundances of the different spruce species. Forest communities in the study area are distributed along geologically based environmental gradients and are shaped by secondary succession following fires and probably floods. Black spruce dominates in the poorly drained, permafrost-rich, and fire-prone backswamp and white spruce in the oppositely characterized meander belt. Although geological chronosequences can be identified along avulsion-prone rivers like the study reach of the Tanana River, superposition of a meander belt-backswamp plan and frequent fire and flood disturbances may negate any vegetation chronosequences older than several centuries.