e18175 Background: EAMs are wearable devices that record aspects of activity, including heart rate and step count. Medical oncologists use performance status (PS) (such as ECOG and Karnofsky) to assess patient’s physical function and to aid treatment decisions. These are subject to patient and physician bias. Our study evaluated the use of EAMs (Fitbit ®) during phase I trials, correlating EAM indices with PS and patient outcomes. Methods: We recruited patients in an academic hospital Drug Development Unit. Patients participating in a phase I trial consented to wear an EAM prior to starting and during the first treatment cycle. Physicians assessed PS at 4 timepoints. Outcomes (progression free survival (PFS), overall survival (OS), adverse events) were recorded. Results: Twenty-five patients were enrolled. At the time of data analysis results were available for 16 patients. Patients wore the EAM for an average of 70.3% (standard deviation 21%) of their time on study, defined as time with recordable heart rate. Mean step count per day varied with ECOG PS although the correlation did not reach statistical significance (co-efficient -1611.1 steps for each change in ECOG PS, p = 0.07, mixed effect regression model). A higher mean step count/day correlated with longer OS (HR0.67, p = 0.01, Cox proportional hazard model). Physician-assessed PS did not correlate with OS (HR2.65, p = 0.37, Cox proportional hazard model). There was no correlation between mean step count and PFS (HR1.7, p = 0.39, Cox proportional hazard model). Conclusions: The use of EAMs is feasible in this patient group. A higher step count correlated with longer OS whereas physician-assessed PS did not. Further research into the potential of EAMs as a means of predicting patient survival to guide decisions on eligibility for phase I trials and the potential for devising a revised Royal Marsden prognostic score incorporating EAM indices is warranted. Clinical trial information: SE536. [Table: see text]