This study investigates platelet-related subtypes in non-small cell lung cancer (NSCLC) and seeks to identify genes associated with prognosis, focusing on the clinical significance of the chloride ion channel gene BEST3. We utilised sequencing and clinical data from GEO, TCGA and the Xena platform, building a risk model based on genetic features. TCGA and GSE37745 served as training cohorts, while GSE50081, GSE13213, GSE30129 and GSE42127 were validation cohorts. Immunotherapy datasets (GSE135222, TCGA-SKCM) were also analysed. Differentially expressed genes (DEGs) were identified using Limma, subtypes through ConsensusClusterPlus and key prognostic genes using COX regression, Random Forest and LASSO-COX. BEST3 expression was validated by flow cytometry (FCM) and functional assays in A549 cells with lentiviral overexpression evaluated its impact on apoptosis, proliferation and migration. Three platelet-related subtypes were identified, with ten key prognostic genes (including BEST3). Gene Ontology (GO) analysis showed six genes involved in platelet pathways. BEST3 was highly expressed in the platelet subtype 1. Flow cytometry confirmed elevated BEST3 levels in NSCLC (35.9% vs. 27.3% in healthy individuals). Overexpression of BEST3 in NSCLC cells suppressed apoptosis and promoted proliferation and migration. The discovery of three platelet subtypes and the role of BEST3 in promoting tumour growth and migration highlights its potential as a therapeutic target and prognostic marker in NSCLC.