9616 Background: Diagnosis of upper limb lymphedema (LE) secondary to treatment for cancer has been complicated by the range of diagnostic thresholds used. Traditional thresholds were arbitrarily chosen but recently normative-based cut-offs have been established. However, the diagnostic power of these thresholds has not been compared to changes identified with lymphoscintigraphy, the gold standard for diagnosis of LE. The aim of this study was to determine thresholds for commonly used clinical diagnostic tools based on lymphatic imaging outcomes. Methods: Women previously diagnosed with LE secondary to treatment for cancer (n=67), and women without LE (n=20) participated. Lymphoscintigraphy was completed and the presence and severity of dermal back flow qualitatively scored as none to mild or moderate to severe by an experienced nuclear medicine physician. On the same day, circumference measurements and segmental bioimpedance spectroscopy (BIS) for 10 cm intervals from the ulnar styloid to 40 cm proximal were recorded. The BIS inter-limb ratio for each segment and the inter-limb circumference differences were compared to diagnostic thresholds based on 2 and 3 standard deviations (SD) above the mean from normative data. Results: The number of BIS segments and circumferences measurements above a 3SD threshold correlated significantly with the dermal backflow score (Rs= 0.710 and 0.824 respectively).The number of abnormal BIS segments detected did not differ significantly between a 2SD and 3SD threshold (χ2=0.23 p=0.63). Of the 20 new abnormal segments detected, 7 were from participants with no or mild evidence of dermal backflow on lymphoscintigraphy including 3 from control participants. In contrast, 46 new abnormal inter-limb circumference differences, all from those with a LE diagnosis, were detected using a 2SD threshold (χ2= 31.785, p<0.001) Conclusions: The need for a standardized, evidenced-based approach for identification of LE is essential. We recommend diagnostic thresholds for segmental BIS be set at 3SD above the mean to minimize false positive diagnoses whereas a lower threshold of 2SD is necessary for the less sensitive inter-limb circumference difference measurements.