The period from 1979 to 1991 witnessed the emergence of the propensity score as a practical tool for tackling time-varying confounding in observational studies. The shift from high‑dimensional covariate adjustment to a scalar balancing score enabled straightforward matching, stratification, and weighting to reduce bias. Researchers focused on formalizing balancing properties and developing time-varying implementations that support causal inference in longitudinal settings. Historical Significance: This era established the propensity score as a unifying framework for causal inference in nonrandomized data, with lasting influence on longitudinal analyses and time-varying treatment evaluation. Key breakthroughs included demonstrations that propensity score methods can balance covariates across treatment groups, effective strategies for subclassification and multivariate matching, and the introduction of sensitivity analyses for unobserved confounders. The legacy of these developments endures in modern observational analysis and policy evaluation.