Abstract only Introduction: Paroxysmal afib is the most common cause of cryptogenic strokes. Long-term insertable cardiac monitor (ICM) has been increasingly used for ongoing evaluation of afib in these patients. Due to financial affordability, it is important to identify appropriate candidates for its routine use. Methods: We reviewed EMRs of a cohort of cryptogenic stroke patients from February 2014 to May 2017 who received REVEAL LINQ ICMs. All patients met ESUS criteria prior to insertion and were enrolled in remote follow-up. At least one year monitoring was conducted if no afib was recorded. Patient demographics, stroke characteristics and risk factors were compared between patients with and without afib. Results: Among total 348 patients enrolled, 99 (28.5%) were found to have afib with median time to afib detection of 128 days. For patients without afib, the median length of follow up was 566 days. Patients with afib were significantly older (mean ± SD, 73.0 ± 9.42 vs. 64.4 ± 11.5, p<0.00001) and their left atrium size (mm) were significantly larger (37.9 ± 7.26 vs. 35.4 ± 5.81, p<0.005) than those patients without afib. Non-afib patients had statistically more frequent association with ongoing smoking or LDL ≥ 120 comparing with afib patients (p<0.05). Although not statistically significant, there was tendency that posterior circulation only strokes occurred less in afib group (p=0.08). However, intra- and extracranial atherosclerosis were comparable between afib and non-afib groups. Within the afib group, afib was detected within 30 days in 25 patients with median time to detection of 14 days. These patients had significant higher rate of intracranial stenosis (p=0.01) comparing with those patients with afib detected beyond 30 days. Conclusion: Paroxysmal afib related strokes account for about one-fourth of cryptogenic strokes in our cohort. Long term ICM may benefit more in older patients with enlarged left atrium and lack of risk factors such as smoking and high LDL. It likely has low yield in patients with posterior infarcts only. Noninvasive ambulatory ECG monitoring for 30 days may be considered first in patients with intracranial stenosis.