The combination of microvascular invasion and PSA levels before salvage radiation therapy (SRT) may provide a better way to stratify the risk of biochemical recurrence after SRT. Abstract Objectives The aim of this study was to identify risk factors to predict a biochemical recurrence (BCR) in patients treated with salvage radiation therapy (SRT) after radical prostatectomy (RP). Methods We retrospectively reviewed 122 Japanese patients who received SRT for BCR after RP. Using uni- and multivariate Cox proportional hazard models, we identified the predictive factors of BCR after SRT. Results With a median follow-up of 61.3 months, 45.9% of the patients showed BCR after SRT, with 61.5 and 41.8% of non-BCR rates at the second and fifth years. Univariate proportional hazards analysis demonstrated that extraprostatic disease (P = 0.029), seminal vesicle invasion (P = 0.005), microvascular invasion (P = 0.001), postoperative Gleason score (P = 0.008) and pre-SRT prostate-specific antigen (PSA) (P = 0.005) were significantly associated with BCR after SRT. However, only the presence of microvascular invasion and a higher pre-SRT PSA were significant predictors in the multivariate analysis. The non-BCR rate in the second year after SRT for 15 patients with microvascular invasion and pre-SRT PSA > 1.2 ng/ml was only 21% compared to 72.5% of 72 patients with negative microvascular invasion and a pre-SRT PSA of <1.2 ng/ml (P = 0.000031). Conclusions While SRT is the most important secondary treatment option for patients with BCR after RP, the effectiveness of SRT may not be uniform. The combination of risk factors such as microvascular invasion in RP specimens and pre-SRT PSA may provide a better way to stratify the risk of BCR after SRT.