Although fibroblast growth factor-2 (FGF-2) participates in the response to vascular injury, the role of cellular deformation in FGF-2 release is incompletely understood. To test the hypothesis that mechanical strain tightly controls FGF-2 release, a novel device was used to impose homogeneous and uniform biaxial strain to human vascular smooth muscle cells. Release of FGF-2 increased with the number of cycles of strain (14%, 1 Hz); 1, 9, and 90 cycles of strain, respectively, released 0.55±0.06%, 2.9±0.3%, and 5.5±1.3% of the total cellular FGF-2 (versus 0.00±0.40% for control, P<.05), but release was not further increased for strain of 90 to 90 000 cycles. Mechanical release of FGF-2 depended on both the frequency and amplitude of deformation. For example, strain (90 cycles, 1 Hz) at 4% amplitude released only 0.1±0.1% of the total FGF-2, but strain at 14% and 33% amplitudes, respectively, released 5.7±0.5% and 19.0±3.0% of the FGF-2 cellular pool (P<.05), suggesting a strain amplitude threshold for FGF-2 release. Injury to a subpopulation of cells increased with the frequency and amplitude of strain, but cells were not injured by strains below 10% amplitude. Strain following pretreatment with heparin released 12.6±1.6% of the total FGF-2 (versus 15.8±0.9% for strain alone, P<.05), indicating that most FGF-2 was liberated from the nuclear or cytoplasmic pools and not from low-affinity extracellular receptors. Conversely, strain in the presence of heparin released 25.2±3.5% of the total FGF-2 (versus 15.6±2.6% for strain alone, P<.05). Thus, cellular strain closely modulates the release of intracellular FGF-2 from human vascular smooth muscle cells, but FGF-2 release is negligible in response to the smaller strains that occur in the normal artery. In addition, larger mechanical strains lead to transfer of intracellular FGF-2 to the extracellular low-affinity receptors, where FGF-2 may be displaced by heparin. These observations provide insight into the mechanisms by which deforming vascular injury, such as that produced by arterial interventions, may elicit a proliferative response.