Background: Brain or liver metastases are the main causes of lung cancer-related death. Predicting liver or brain metastases before having evidence from imaging of the tumors is challenging, but important for early patient intervention. Accumulating evidence has suggested that circulating exosomes in the blood could mediate cancer metastasis by transferring specific proteins to the target cells. Methods: Using liquid chromatography-MS/MS, we analyzed the proteomic profiles of plasma exosomes derived from 42 metastatic lung cancer patients (including 26 solitary brain metastasis (BM) and 16 solitary liver metastasis (LM)), 25 local advanced (LA) lung cancer patients without metastasis, and 5 healthy controls (HC), and explored organ-specific proteomic biomarkers. Results: 120 and 143 differentially expressed exosomal proteins were found to be the dysregulated candidates in BM and LM of lung cancer (BM-DEEPs, LM-DEEPs), as compared to the LA lung cancer samples, respectively. The bioinformatics analyses indicated the homogeneity and heterogeneity in BM-DEEPs and LM-DEEPs. In terms of homogeneity, they were mainly involved in the protein activation cascade, collagen-containing extracellular matrix, and ECM-receptor interaction. In terms of heterogeneity, the BM-DEEPs were significantly enriched in S100 protein and calcium-dependent protein, while the LM-DEEPs were mainly enriched in integrin, lipoprotein, heat shock protein and proteoglycans. Moreover, our study also revealed the heterogeneity in plasma-derived exosome proteomics in patients with non-small cell lung carcinoma (NSCLC) and small cell lung carcinoma (SCLC), which partly explained the different metastatic characteristics of NSCLC and SCLC. Additionally, we found that MUC5B and SELL could be used as a reliable diagnostic marker of BM, while APOH, CD81 and CCT5 could help to diagnose LM in lung cancer patients. Conclusions: For the first time, we revealed the comprehensive and comparative proteomic profiles of plasma-derived exosomes from lung cancer patients with solitary brain and liver metastasis. We further proposed some potential biomarkers and pathological mechanisms, which could serve as excellent resource for further studies investigating lung cancer metastasis.