Senescence in cancer cells acts as a tumor suppressor, whereas in fibroblasts enhances tumor growth. Senescence has been reported in tumor associated fibroblasts (TAFs) from a growing list of cancer subtypes. However, the presence of senescent TAFs in lung cancer remains undefined. We examined senescence in TAFs from primary lung cancer and paired control fibroblasts from unaffected tissue in three major histologic subtypes: adenocarcinoma (ADC), squamous cell carcinoma (SCC) and large cell carcinoma (LCC). Three independent senescence markers (senescence-associated beta-galactosidase, permanent growth arrest and spreading) were consistently observed in cultured LCC-TAFs only, revealing a selective premature senescence. Intriguingly, SCC-TAFs exhibited a poor growth response in the absence of senescence markers, indicating a dysfunctional phenotype rather than senescence. Co-culturing normal fibroblasts with LCC (but not ADC or SCC) cancer cells was sufficient to render fibroblasts senescent through oxidative stress, indicating that senescence in LCC-TAFs is driven by heterotypic signaling. In addition, senescent fibroblasts provided selective growth and invasive advantages to LCC cells in culture compared to normal fibroblasts. Likewise, senescent fibroblasts enhanced tumor growth and lung dissemination of tumor cells when co-injected with LCC cells in nude mice beyond the effects induced by control fibroblasts. These results define the subtype-specific aberrant phenotypes of lung TAFs, thereby challenging the common assumption that lung TAFs are a heterogeneous myofibroblast-like cell population regardless of their subtype. Importantly, because LCC often distinguishes itself in the clinic by its aggressive nature, we argue that senescent TAFs may contribute to the selective aggressive behavior of LCC tumors.