In low and middle-income countries, estimating the proportion of vaccinated toddlers in a population is important for controlling vaccine-preventable diseases by identifying districts where immunization services need strengthening. Estimates measured before and several years after specific interventions can assess program performance. However, employing different methods to derive vaccination coverage estimates often yield differing results. METHODS: Linked vaccination coverage surveys and seroprotection surveys performed among ~300 toddlers 12-23 months of age in districts (woredas), one per region, of Ethiopia (total, ~900 toddlers) in 2013 to estimate the proportion vaccinated with tetanus toxoid (a proxy for pentavalent vaccine) and measles vaccine. The surveys were followed by implementation of the Reaching Every District using Quality Improvement (RED-QI) approach to strengthen the immunization system. Linked coverage/serosurveys were repeated in 2016 to assess effects of the interventions on vaccination coverage. Indicators included "documented coverage" (vaccination card and/or health facility register records) and "crude coverage" (documented plus parent/caretaker recall for children without cards). Seroprotection thresholds were IgG-ELISA tetanus antitoxin ≥0.05 IU/ml and plaque reduction neutralization (PRN) measles titers ≥120 mIU/ml. FINDINGS: Improved markers in 2016 over 2013 include coverage of pentavalent vaccination, vaccination timeliness, and fewer missed opportunities to vaccinate. In parallel, tetanus seroprotection increased in the 3 woredas from 59.6% to 79.1%, 72.9% to 83.7%, and 94.3 to 99.3%. In 2015, the Ethiopian government conducted supplemental measles mass vaccination campaigns in several regions including one that involved a project woreda and the campaign overlapped with the RED-QI intervention timeframe; protective measles PRN titers there rose from 31.0% to 50.0%. INTERPRETATION: The prevalence of seroprotective titers of tetanus antitoxin (stimulated by tetanus toxoid components within pentavalent vaccine) provides a reliable biomarker to identify children who received pentavalent vaccine. In the three study woredas, the RED-QI intervention appeared to improve immunization service delivery, as documented by enhanced pentavalent vaccine coverage, vaccination timeliness, and fewer missed vaccination opportunities. A measles mass vaccination campaign was followed by a markedly increased prevalence of measles PRN antibodies. Collectively, these observations suggest that wider implementation of RED-QI can strengthen immunization, and periodic linked vaccination surveys/serosurveys can monitor changes.