This longitudinal study explored the impact of COVID‐19 on individuals in romantic relationships. The sample spans three waves: first confinement (n = 52), deconfinement (n = 49), and second confinement (n = 26). The study tested sociodemographic factors, psychological adjustment (anxiety, depression, stress, well‐being), COVID‐19 threat perception, dyadic coping, and relationship quality. Results from repeated measures ANOVA and fuzzy set qualitative comparative analysis (fsQCA) models revealed a decline in anxiety, depression, and stress symptoms, coupled with improved relationship quality over time. Well‐being and dyadic coping remained stable, while COVID‐19 threat perception increased. QCA models emphasized the predictive power of initial mental states (anxiety, depression, stress, and well‐being at W1) on subsequent adjustment. Notably, shorter relationship duration, healthcare worker status, and not having children, when combined with prior mental states, explained increased symptoms and diminished well‐being. The study underscores the significance of addressing these factors in individuals navigating romantic relationships during the pandemic.