OBJECTIVES: To characterize stepping in a nationally representative sample of U.S. adults by age, sex, and race and ethnicity. METHODS: Step counts are a simple, scalable metric of physical activity associated with lower mortality and chronic disease risks. However, nationally representative estimates of step volume and intensity across demographic groups remain limited. We analyzed data from U.S. adults aged ≥20 years in NHANES 2011-2014 who wore a wrist accelerometer for seven days. Stepping metrics derived from the OxWearables machine-learning algorithm, including daily steps, peak 30-minute cadence, and self-selected cadence. We estimated survey-weighted means and the prevalence of meeting age-specific thresholds associated with substantial reductions in mortality risk (≥9,000 steps/day for adults <60 years; ≥7,000 steps/day for adults ≥60 years) in previous meta-analyses. RESULTS: Adults (N=8,762, mean age 48 years [SD 17]; 52% female) averaged 8,818 steps/day (95% CI: 8,676-8,961). 51% of steps were accumulated <2-minute bouts and 28% from ≥5-minute bouts. Males took more steps than females. Hispanics had higher step counts than other racial and ethnic subgroups. Peak and self-selected cadence converged in older adults. 48.6% of adults <60 years obtained ≥9,000 steps/day, and 44.8% of ≥60 years obtained ≥7,000 steps/day. CONCLUSIONS: Nearly half of U.S. adults achieved step counts associated with substantially lower mortality risk, with differences by age, sex, race, and ethnic groups. We provide benchmarks for population surveillance and can inform strategies to promote step-based activity. These findings could guide future efforts to promote stepping as physical activity among the least active adults.