AbstractBackgroundThe workplace is a setting for severe acute respiratory syndrome coronavirus 2 (SARS‐CoV‐2) transmission via respiratory droplets and aerosols for those working in close proximity to others. Currently, limited methods exist for assessing SARS‐CoV‐2 exposure. Since occupation serves as a surrogate measure, job exposure matrices (JEMs) can be useful for population‐based exposure assessment for a portion of exposure.MethodsWe developed a JEM to assess physical proximity at work. Scores for questions related to frequency of face‐to‐face discussions, working closely with others, and working with a team were extracted from a US‐based, comprehensive source of descriptive occupational information (Occupational Information Network [O*NET]). We described score distributions using univariate analyses, devised thresholds, and assigned exposure levels for 968 O*NET occupations. Three exposure measures were constructed using combinations of O*NET data, with expert judgment, and accounting for telework. National and California employment estimates were used to assess the workforce proportions by proximity level and demographic characteristics.ResultsWe categorized 535 US Census occupations (2010) into four‐level ordinal exposure levels (not close to very close). Overall, an estimated 56% of the California workforce worked in very close proximity, which decreased to 46% when accounting for telework. The occupational groups working very close across all three measures were: healthcare support, healthcare practitioner, food preparation and serving, building and grounds cleaning and maintenance, and protective service occupations. Latinos and women were overrepresented within occupations working in very close physical proximity.ConclusionJEMs can inform SARS‐CoV‐2 exposure assessment for epidemiologic studies, assist in resource allocation, and inform prevention strategies.