Despite being nearly universally recognised as a virtue, judicial independence has been challenged in almost all parts of the world. Some commentators even consider it to be so open to differing interpretations as to be a useless concept, that should be unpacked to its smaller components to be studied meaningfully. We are less cynical about the idea. According to our theory, judicial independence exists where powerful actors are unable or unwilling to inappropriately interfere with the workings of the judiciary. Judicial independence is thus a relational concept and always results from the interplay between the capacity and willingness of powerful actors to inappropriately interfere with the judiciary, and the capacity and willingness of judicial actors and their allies to withstand such actions. We distinguish three levels of judicial independence: de jure institutional independence, de facto institutional independence, and decisional independence. Courts are thus independent when powerful actors do not consistently impose their preferences in disputes they have a stake in, either by capturing the courts through formal changes of laws governing the judiciary, through rigging these laws in their favour, or by skewing judicial decision-making. By contrast, a dependent judiciary is the one that is captured, rigged, or skewed.