In the early twentieth century, there was no consensus about the nature of climate. Focusing on ecological understandings of climate, this article contends that the diverse scientific practices used by ecologists to reconstruct past climates resulted in different notions of climate. At the turn of the century, a number of ecologists looked to the geological past to understand climate, but they found different climatic patterns depending on the way they encountered it. Some ecologists, led by Frederic E. Clements, came to believe that the climate displayed cyclical, recurrent patterns after examining tree rings. Others, notably Paul B. Sears, believed that the climate was characterized by unpredictable complexity after examinations of fossil pollen. Their divergent understandings of climate came to a head during the Dust Bowl of the 1930s, a severe drought whose effects were amplified by human actions. Both ecologists provided advice to governments and farmers based on their notions of whether or not climate displayed regular periodicity. Their varied assessments of climate reveal broader ideas about interwar science and culture, especially contestations over the capacity of science to predict the future.