Cross-sectional studies have suggested a role for activation of adaptive immunity in smokers with emphysema, but the clinical application of these findings has not been explored. Here we examined the utility of detecting autoreactive T cells as a screening tool for emphysema in an at-risk population of smokers. We followed 156 former and current (ever)-smokers for 2 years to assess whether peripheral blood CD4 T cell cytokine responses to lung elastin fragments (EFs) could discriminate between those with and without emphysema, and to evaluate the relevance of autoreactive T cells to predict changes during follow-up in lung physiological parameters. Volunteers underwent baseline complete phenotypic assessment with pulmonary function tests, quantitative chest CT, yearly 6-min walk distance (6MWD) testing, and annual measurement of CD4 T cell cytokine responses to EFs. The areas under the receiver operating characteristic curve to predict emphysema for interferon gamma (IFN-γ), and interleukin 6 (IL-6) responses to EFs were 0.81 (95% CI of 0.74–0.88) and 0.79 (95% CI of 0.72–0.86) respectively. We developed a dual cytokine enzyme-linked immunocell spot assay, the γ-6 Spot, using CD4 T cell IFN-γ and IL-6 responses and found that it discriminated emphysema with 90% sensitivity. After adjusting for potential confounders, the presence of autoreactive T cells was predictive of a decrease in 6MWD over 2 years (decline in 6MWD, −19 m per fold change in IFN-γ; P = 0.026, and −26 m per fold change in IL-6; P = 0.003). In support of the human association studies, we cloned CD4 T cells with characteristic T helper (Th)1 and Th17 responses to EFs in the peripheral blood of ever-smokers with emphysema, confirming antigenicity of lung elastin in this population. These findings collectively suggest that the EF-specific autoreactive CD4 T cell assay, γ-6 Spot, could provide a non-invasive diagnostic tool with potential application to large-scale screening to discriminate emphysema in ever-smokers, and predict early relevant physiological outcomes in those at risk.