Individualism–collectivism (I-C), the most widely researched cultural dimension, is often equated with Hofstede’s pioneering nation scores. Concerns are growing about these scores’ validity, but subsequent research has not produced a widely accepted alternative. Here, we offer a refined theoretical conceptualization of I-C, we systematically reevaluate the validity of Hofstede’s I-C scores, and we report the development and validation of a new I-C index, covering 102 countries/territories inhabited by an estimated 88% of the world’s population. In Study 1, we document the inferior convergent and nomological validity of Hofstede’s I-C index, compared to subsequent measures. Hofstede’s scores substantially overestimate individualism in English-speaking countries and collectivism in East Asian societies, which we demonstrate can considerably bias research findings. In Study 2, we develop an authoritative, theory-driven I-C index, using nationally representative data from the World Values Survey and European Values Study, which shows excellent internal coherence, temporal stability, and strong evidence of convergent, discriminant, and nomological validity. Theorized facets of individualism—freedom (vs. conformity), tolerance of differences (vs. exclusion), and equality (vs. discrimination)—form a coherent and stable dimension at the cultural level of analysis. Individualism is higher in societies with better existential security (e.g., socioeconomic development, stable institutions) and is not associated with greater selfishness, anomie, or competitive beliefs and values. Relying on outdated indices may perpetuate cultural stereotypes and underpin flawed theorizing. Scholars should use theoretically appropriate and up-to-date measures of societal culture when seeking to understand global variation in human psychological functioning.