Evaluate various indirect predictors of peak skin dose during fluoroscopically-guided interventional procedures to find the best marker to identify patients beyond thresholds for skin injury.High risk (higher dose) procedures including embolization, TIPS, visceral Tc-MAA, and renal/mesenteric angiography and stent placement procedures were selected for analysis. Procedural dose protocol pages displaying tube angles and respective doses were obtained weekly from September 2011 to August 2012. Peak skin doses (PSD) were estimated using techniques developed at our institution for cases that exceed the 60 minute fluoroscopy threshold per our radiation safety protocol. Indirect predictors of peak skin dose were assessed and the best cutoff values were selected to predict 2 Gray (Gy) and 3 Gy PSDs as minimum thresholds for possible skin damage. Statistical analysis was performed to compare the selected best cutoffs to values quoted in recent literature.A total of 208 procedures were collected. The mean PSD (1.5±1.1 Gy), mean total dose (2.2±1.6 Gy) as total air kerma (TAK), mean total fluoroscopy time (FT) (25±18.6 min), mean total dose-area-product (DAP) (423±302 Gycm2), and mean total procedure time (212±79.6 min) were calculated. FT (30/60 min) was a very poor predictor of PSD, statistically underperforming compared to TAK and DAP at all levels. At the 2 Gy PSD threshold, TAK (2.7 and 3.0 Gy) performed slightly better than DAP (437 and 300 Gycm2) with our best cutoffs not statistically different from reported cutoffs. At the 3 Gy PSD threshold, our best TAK cutoff (3.9 Gy) significantly outperformed the reported 5 Gy TAK cutoff as well as DAP (647 and 500 Gycm2) equivalents which were similar to each other. Specificity, sensitivity, NPV, PPV, accuracy, and confidence interval analysis was performed.Fluoroscopy time was a very poor predictor of PSD during interventional procedures and statistically underperformed compared to total dose (TAK). Doses of 2.7 Gy and 3.0 Gy (TAK) were statistically similar to predict a PSD of 2 Gy, while our cutoff of 3.9 Gy TAK had a significantly better overall performance for predicting a PSD of 3 Gy or higher.