BACKGROUND: Our aim was to identify a dose-volume response relationship for brain metastases treated with frameless stereotactic radiosurgery (SRS). METHODS: We reviewed patients who underwent frameless single-fraction linear accelerator SRS for brain metastases between 2007 and 2013 from an institutional database. Proportional hazards modeling was used to identify predictors of outcome. A ratio of maximum lesion dose per mm-diameter (Gy/mm) was constructed to establish a dose-volume relationship. RESULTS: There were 316 metastases evaluated in 121 patients (2 - 33 mm in the largest diameter). The median peripheral dose was 18.0 Gy (range: 10.0 - 24.0 Gy). Local control was 84.8% for all lesions and was affected by location, peripheral dose, maximum dose, and lesion size (p values < 0.050). A dose-volume response relationship was constructed using the maximum dose and lesion size. A unit increase in Gy/mm was associated with decreased local failure (p = 0.005). Local control of 80%, 85%, and 90% corresponded to maximum doses per millimeter of 1.67 Gy/mm, 2.86 Gy/mm, and 4.4 Gy/mm, respectively. Toxicity was uncommon and only 1.0% of lesions developed radionecrosis requiring surgery. CONCLUSIONS: For brain metastases less than 3 cm, a dose-volume response relationship exists between maximum radiosurgical dose and lesion size, which is predictive of local control.