INTRODUCTION: Population health surveys inform and demonstrate the impact of public health policies. However, the performance of such surveys in specific groups of interest (e.g., Hispanics/Latinos in a neighborhood of New York City) is rarely studied.

METHOD: We compared measures for obesity, hypertension, diabetes, and current smoking based on the New York City Community Health Survey (CHS, a telephone survey of New York City adults) with the Hispanic Community Health Survey/Study of Latinos (HCHS/SOL), an in-person survey of Hispanic/Latino adults in four communities in the United States (2008-2011), including the Bronx. CHS data were limited to Hispanic/Latinos living in the HCHS/SOL Bronx catchment area.

RESULTS: Compared with CHS, HCHS/SOL estimated higher prevalence of obesity (in HCHS/SOL, P = 45.0% vs. in CHS, P = 30.6%, .01) and current smoking (P = 21.2% vs. P = 16.2%, .01) but similar for hypertension (P = 33.1% vs. P = 33.8%, > .05) and diabetes (P = 15.2% vs. P = 15.7%, > .05). Stratified estimates (by age, sex, education, and Hispanic/Latino heritage) followed similar trends.

CONCLUSION: Our study emphasizes the importance of assessing potential bias in population-based surveys of Hispanics/Latinos and other populations of interest and highlights the complex nature of measuring health outcomes via population-based surveys.