Cellular senescence is a terminal cell proliferation arrest that can be triggered by oncogenes. One of the traits of oncogene-induced senescence (OIS) is the so-called senescence-associated secretory phenotype or senescence secretome. Depending on the context, the non-cell autonomous effects of OIS may vary from tumor suppression to promotion of metastasis. Despite being such a physiological and pathologically relevant effector, the mechanisms of generation of the senescence secretome are largely unknown. We analyzed by label-free proteomics the secretome of p95HER2-induced senescent cells and compared the levels of the membrane-anchored proteins with their transcript levels. Then, protein and RNA levels of ADAM17 were evaluated by using Western blot and reverse transcription-polymerase chain reaction, its localization by using biotin labeling and immunofluorescence, and its activity by using alkaline phosphatase-tagged substrates. The p95HER2-expressing cell lines, senescent MCF7 and proliferating MCF10A, were analyzed to study ADAM17 regulation. Finally, we knocked down ADAM17 to determine its contribution to the senescence-associated secretome. The effect of this secretome was evaluated in migration assays in vitro and in nude mice by assessing the metastatic ability of orthotopically co-injected non-senescent cells. Using breast cancer cells expressing p95HER2, a constitutively active fragment of the proto-oncogene HER2 that induces OIS, we show that the extracellular domains of a variety of membrane-bound proteins form part of the senescence secretome. We determine that these proteins are regulated transcriptionally and, in addition, that their shedding is limited by the protease ADAM17. The activity of the sheddase is constrained, at least in part, by the accumulation of cellular cholesterol. The blockade of ADAM17 abrogates several prometastatic effects of the p95HER2-induced senescence secretome, both in vitro and in vivo. Considering these findings, we conclude that ectodomain shedding is tightly regulated in oncogene-induced senescent cells by integrating transcription of the shedding substrates with limiting ADAM17 activity. The remaining activity of ADAM17 contributes to the non-cell autonomous protumorigenic effects of p95HER2-induced senescent cells. Because ADAM17 is druggable, these results represent an approximation to the pharmacological regulation of the senescence secretome. The online version of this article (doi:10.1186/s13058-015-0619-7) contains supplementary material, which is available to authorized users