OBJECTIVES: To document variations in time horizons used in published cost-effectiveness analyses (CEAs) that assume a US payer perspective, and examine how the assumed time horizon influences results. METHODS: We systematically reviewed the Tufts Medical Center Cost-Effectiveness Analysis (CEA) Registry. We included all US-based studies that: used a healthcare payer perspective, were published during 2005-2014, and received a Registry quality score of 4 or better on a seven-point scale. We classified the identified CEAs as short-term (time horizon ≤ 5 years) and long-term (> 5 years). We analyzed associations between study characteristics (e.g., intervention type, prevention stage, and funding sources) and the specified time horizon. Finally, we developed case studies with selected interventions to illustrate the association between the time horizon and the cost-effectiveness estimates. RESULTS: The literature review identified 782 articles that met our inclusion criteria. Among the identified studies, 552 studies (74%) utilized a long-term time horizon while 198 studies (25%) used a short-term horizon. Of 32 studies that employed multiple time horizons, we reviewed 23 intervention-specific incremental cost-effectiveness ratios (ICERs). The extension of the time horizon yielded more favorable ICERs in 19 cases, and less favorable ICERs in 4 cases. In our case studies, the use of a longer time horizon also yielded more favorable cost-effectiveness ratios. CONCLUSIONS: The time horizon used in CEAs can substantially influence the value assessments of medical interventions. To capture broad societal impacts on costs and health effects, we encourage the use of time horizons that extend sufficiently into the future.