BACKGROUND The exact mechanisms of lung adenocarcinoma (LUAD) etiology and pathogenesis remain unclear. MATERIAL AND METHODS In this study, we evaluated alternative splicing (AS) events in LUAD. We separately analyzed LUAD data from The Cancer Genome Atlas (TCGA) database and AS-related feature lists from SpliceSeq to develop risk models for AS events and further validated risk models for AS events. The association between immune-related features and the risk model of AS events was evaluated. RESULTS We found that exon skip (ES) and mutually exclusive exons (ME) exhibited the most and least AS events, respectively. The risk score and stage of LUAD patients were strongly associated with prognosis and were independent influences on the prognosis of LUAD. The pathological stage (stage, T, N) and risk model were incorporated to construct a column line graph with better predictive ability. Risk models were associated with tumor microenvironment, and higher risk score values were associated with higher M2 macrophage content and lower M0 macrophage and helper T lymphocyte content. The correlation between core genes and immunity was further assessed by analyzing differentially-expressed genes between risk models. These results suggest an association between the level of risk for AS events and the density of immune cell infiltration. CONCLUSIONS Our findings suggest a clear association between AS risk model and patient prognosis, and was performed to confirm the biological relationship between AS and immunity.