(1) Introduction: The aim of this study was to identify the plasma extracellular vesicle (EV)-specific transcriptional profile in small-cell lung cancer (SCLC) and to explore the application value of plasma EV long RNA (exLR) in SCLC treatment prediction and diagnosis. (2) Methods: Plasma samples were collected from 57 SCLC treatment-naive patients, 104 non-small-cell lung cancer (NSCLC) patients and 59 healthy participants. The SCLC patients were divided into chemo-sensitive and chemo-refractory groups based on the therapeutic effects. The exLR profiles of the plasma samples were analyzed by high-throughput sequencing. Bioinformatics approaches were used to investigate the differentially expressed exLRs and their biofunctions. Finally, a t-signature was constructed using logistic regression for SCLC treatment prediction and diagnosis. (3) Results: We obtained 220 plasma exLRs profiles in all the participants. Totals of 5787 and 1207 differentially expressed exLRs were identified between SCLC/healthy controls, between the chemo-sensitive/chemo-refractory groups, respectively. Furthermore, we constructed a t-signature that comprised ten exLRs, including EPCAM, CCNE2, CDC6, KRT8, LAMB1, CALB2, STMN1, UCHL1, HOXB7 and CDCA7, for SCLC treatment prediction and diagnosis. The exLR t-score effectively distinguished the chemo-sensitive from the chemo-refractory group ( = 9.268 × 10) with an area under the receiver operating characteristic curve (AUC) of 0.9091 (95% CI: 0.837 to 0.9811) and distinguished SCLC from healthy controls (AUC: 0.9643; 95% CI: 0.9256-1) and NSCLC (AUC: 0.721; 95% CI: 0.6384-0.8036). (4) Conclusions: This study firstly characterized the plasma exLR profiles of SCLC patients and verified the feasibility and value of identifying biomarkers based on exLR profiles in SCLC diagnosis and treatment prediction.