Although outcome following bevacizumab among recurrent grade IV malignant glioma patients is documented as poor by several analyses, outcome for recurrent grade III patients following bevacizumab therapy has not been specifically evaluated. We performed a pooled analysis of 96 recurrent grade III malignant glioma patients enrolled on three consecutive phase II bevacizumab salvage trials to evaluate overall outcome following bevacizumab trial discontinuation. Outcome on the three bevacizumab trials, which included similar eligibility, treatment and assessment criteria, was comparable. Forty-nine patients who progressed on bevacizumab trial therapy and remained alive for at least 30 days elected to receive additional therapy. These patients achieved a median PFS-6 and OS of 30.6% (95% CI: 18.4, 43.6) and 10.3 months (95% CI: 5.2, 11.7), respectively. Among patients who continued bevacizumab therapy (n = 23) after study progression, PFS-6 and median OS were 39.1% (95% CI: 19.9, 58.0) and 9.2 months (95% CI: 5.2, 13.6), respectively, compared to 23.1% (95% CI: 9.4, 40.3; P = 0.51) and 10.3 months (95% CI: 2.5, 14.4; P = 0.91) for patients who initiated non-bevacizumab containing therapy (n = 26). Outcome after discontinuation of bevacizumab therapy for recurrent grade III malignant glioma patients is associated with improved outcome compared to historical data for recurrent grade IV malignant glioma patients. Salvage therapies following bevacizumab failure have modest activity for grade III malignant glioma patients that is independent of further bevacizumab continuation.