Purpose This work aimed to study postpartum mental outcomes and determinants of the intake of caffeinated beverages during the pandemic in women from Argentina. Methods This cross-sectional study recruited 619 women who responded to online self-report questionnaires during the first and second waves of COVID-19, including validated instruments (Insomnia Severity Index, Perceived Stress Scale, Postpartum Depression Screening Scale, Memory Complaint Scale, and Breastfeeding Self-Efficacy Scale), and general data. Intake frequency and amount of caffeinated beverages were estimated. Multivariate regression and structural equation models identified associations and effects (p < 0.05). Results Women were under social restrictions for 60.39 days, with home and essential activities increasing caffeinated intake. They ingested (mL/d): yerba mate (1457.71), coffee (66.85), tea (67.61), and soft drinks (50.95), which provided 646.20 mg/d of caffeine. Intakes of coffee and yerba mate were higher than pre-pandemic ones. Coffee was positively associated with stress and insomnia, and indirectly linked to higher levels of depression and memory complaints, and lower breastfeeding self-efficacy. Tea showed a similar but weaker association. Yerba mate correlated inversely with depression (through direct pathways), insomnia, and memory complaints (through indirect pathways), promoting breastfeeding self-efficacy. Soft drinks and caffeine did not present significant associations. Conclusion Although findings do not imply causation, results suggest that beverages would exhibit caffeine-independent affective and cognitive roles, which might be anxiogenic in the case of coffee and tea (to a lesser extent). Yerba mate showed antidepressant potential. Given that breastfeeding might be compromised during the pandemic, yerba mate intake is promissory to protect postpartum mental health.