Background Findings from published studies suggest that exposure to and interactions with green spaces are associated with improved psychological wellbeing and have cognitive, physiological, and social benefits, but few studies have examined their potential effect on the risk of mortality. We therefore undertook a national study in Canada to examine associations between urban greenness and cause-specific mortality. Methods We used data from a large cohort study (the 2001 Canadian Census Health and Environment Cohort [2001 CanCHEC]), which consisted of approximately 1·3 million adult (aged ≥19 years), non-immigrant, urban Canadians in 30 cities who responded to the mandatory 2001 Statistics Canada long-form census. The cohort has been linked by Statistics Canada to the Canadian mortality database and to annual income tax filings through 2011. We measured greenness with images from the moderate-resolution imaging spectroradiometer from NASA's Aqua satellite. We assigned estimates of exposure to greenness derived from remotely sensed Normalized Difference Vegetation Index (NDVI) within both 250 m and 500 m of participants' residences for each year during 11 years of follow-up (between 2001 and 2011). We used Cox proportional hazards models to estimate associations between residential greenness (as a continuous variable) and mortality. We estimated hazard ratios (HRs) and corresponding 95% CIs per IQR (0·15) increase in NDVI adjusted for personal (eg, education and income) and contextual covariates, including exposures to fine particulate matter, ozone, and nitrogen dioxide. We also considered effect modification by selected personal covariates (age, sex, household income adequacy quintiles, highest level of education, and marital status). Findings Our cohort consisted of approximately 1 265 000 million individuals at baseline who contributed 11 523 770 person-years. We showed significant decreased risks of mortality in the range of 8–12% from all causes of death examined with increased greenness around participants' residence. In the fully adjusted analyses, the risk was significantly decreased for all causes of death (non-accidental HR 0·915, 95% CI 0·905–0·924; cardiovascular plus diabetes 0·911, 0·895–0·928; cardiovascular 0·911, 0·894–0·928; ischaemic heart disease 0·904, 0·882–0·927; cerebrovascular 0·942, 0·902–0·983; and respiratory 0·899, 0·869–0·930). Greenness associations were more protective among men than women (HR 0·880, 95% CI 0·868–0·893 vs 0·955, 0·941–0·969), and among individuals with higher incomes (highest quintile 0·812, 0·791–0·834 vs lowest quintile 0·991, 0·972–1·011) and more education (degree or more 0·816, 0·791–0·842 vs did not complete high school 0·964, 0·950–0·978). Interpretation Increased amounts of residential greenness were associated with reduced risks of dying from several common causes of death among urban Canadians. We identified evidence of inequalities, both in terms of exposures to greenness and mortality risks, by personal socioeconomic status among individuals living in generally similar environments, and with reasonably similar access to health care and other social services. The findings support the development of policies related to creating greener and healthier cities. Funding None. © 2017 The Author(s). Published by Elsevier Ltd. This is an Open Access article under the CC BY-NC-ND 4.0 license

[No Keywords available]

urban population urban area immigrant high school residential home follow up human middle aged diabetes mellitus controlled study data base measurement Aged marriage Young Adult population size suspended particulate matter satellite imagery male Canada particle size female tax environmental exposure population research Ozone Canadian Article major clinical study household income adult mortality age cardiovascular mortality ischemic heart disease cohort analysis nitrogen dioxide cause of death environmental health mortality risk risk reduction sex highest income group population statistics