To evaluate the effect of 0.1%-fluorometholone (FML) on tear inflammatory molecule levels after 22-days treatment in dry eye disease (DED) patients exposed to an adverse controlled environment (ACE), identifying different biomarkers. Analysis of a double-masked randomized clinical trial. Forty-one DED patients received 4-drops daily of topical FML (FML-group) or polyvinyl-alcohol (PA-group) for 22 days. At day 21, patients were exposed to an ACE. Tear samples were collected at V1 (baseline), V2 (pre-ACE), V3 (post-2-h-ACE) and V4 (24-h post-ACE). Concentrations of 18 molecules (EGF, IFN-γ, TNF-α, IL-1β, IL-1RA, IL-2, IL-4, IL-6, IL-8/CXCL8, IL-10, IL-12, IL-13, IL-17A, IP-10/CXCL10, MCP-1/CCL2, MIP-1α/CCL3, RANTES/CCL5 and MMP-9) were analyzed. Similarities among patients in molecule concentrations at V1 were evaluated. A linear-mixed effect model analyzed the influence of different variables on concentrations changes. Multidimensional scaling (MDS) divided patients into two groups based on differences in EGF, IFN-γ, IL-8/CXCL8, RANTES/CCL5, and MMP-9 levels at V1. Groups had different clinical severities based on Schirmer test and conjunctival and corneal staining. IL-1RA, IL-2, and TNF-α were differentially affected by time, depending on treatment. Between V2-V3, there were significant changes in EGF, IL-1RA, IL-2, IL-8/CXCL8, IL-13, IP-10/CXCL10, TNF-α, and MMP-9. The strongest biomarker candidates were IFN-γ, RANTES/CCL5, and MMP-9 as DED severity biomarkers; IL-2 as DED therapeutic biomarker; and EGF as DED activity biomarker. This clinical trial design using a controlled environment and the identified tear biomarkers could be useful to objectively select target patients, to define stress response, and to evaluate therapeutic endpoints in clinical trials.