57 studies from 35 cohorts were included in the systematic review and 31 studies from 24 cohorts were included in meta-analyses. For all-cause mortality, cardiovascular disease incidence, dementia, and falls, an inverse non-linear dose-response association was found, with inflection points at around 5000–7000 steps per day. An inverse linear association was found for cardiovascular disease mortality, cancer incidence, cancer mortality, type 2 diabetes incidence, and depressive symptoms. Based on our meta-analyses, compared with 2000 steps per day, 7000 steps per day was associated with a 47% lower risk of all-cause mortality, a 25% lower risk of cardiovascular disease incidence, a 47% lower risk of cardiovascular disease mortality, a non-significant 6% lower risk of cancer incidence, a 37% lower risk of cancer mortality, a 14% lower risk of type 2 diabetes, a 38% lower risk of dementia, a 22% lower risk of depressive symptoms, and a 28% lower risk of falls. Studies on physical function (not based on meta-analysis) reported similar inverse associations. The evidence certainty was moderate for all outcomes except for cardiovascular disease mortality (low), cancer incidence (low), physical function (low), and falls (very low).