In recent years, researchers have become increasingly interested in exploring frequency as a source of variability in heritage speakers’ (HSs) knowledge of their heritage language (HL). While many of these studies acknowledge that frequency can affect the shape of HL grammars, there is still no clear consensus about (a) what “frequency” means in the context of HL acquisition and (b) how to operationalize its multiple subtypes. In this paper, we provide a critical overview of frequency effects in HL research and their relevance for understanding patterns of inter/intra-speaker variability. To do so, we outline how prior research has defined, measured, and tested frequency, and present—as well as evaluate—novel methodological approaches and innovations recently implemented in the study of frequency effects, including a new analysis of how self-reported lexical frequency reliably predicts HSs’ production of subjunctive mood in Spanish. Our aim is to highlight the immense potential of such work for addressing long-standing questions about HL grammars and to propose new lines of inquiry that will open up additional pathways for understanding HL variability.