Trait based ecology has developed fast in the last decades, aiming to both explain mechanisms of community assembly, and predict patterns in nature, such as the effects of biodiversity shifts on key ecosystem processes. This body of work has stimulated the development of several conceptual frameworks and analytical methods, as well as the production of trait databases covering a growing number of taxa and organizational levels (from individuals to guilds). However, this breeding ground of novel concepts and tools currently lacks a general and coherent framework, under which functional traits can help ecologists organize their research aims, and serve as the common currency to unify several scientific disciplines. Specifically, we see a need to bridge the gaps between community ecology, ecosystem ecology, and evolutionary biology, in order to address the most pressing environmental issues of our time. To achieve this integration goal, we define a trait-integration continuum, which reconciles alternative trait definitions and approaches in ecology. This continuum outlines a coherent progression of biological scales, along which traits interact and hierarchically integrate from genetic information, to whole organism fitness-related traits, to trait syndromes and functional groups. Our conceptual scheme proposes that lower-level trait integration is closer to the inference of ecoevolutionary mechanisms determining population and community properties, whereas higher-level trait integration is most suited to the prediction of ecosystem processes. Within these two extremes, trait integration varies on a continuous scale, which relates directly to the inductive-deductive loop that should characterize the scientific method. With our proposed framework, we aim to facilitate scientists in contextualising their research based on the trait-integration levels that matter most to their specific goals. Explicitly acknowledging the existence of a trait-integration continuum is a promising way for framing the appropriate questions, thus obtaining reliable answers and results that are comparable across studies and disciplines.