Background and objectives Facility-level variation has been reported among veterans receiving care for various diseases. We studied the frequency and facility-level variations of guideline-recommended practices in patients with diabetes and CKD. Design, setting, participants, & measurements Patients with diabetes and concomitant CKD (eGFR 15-59 ml/min per 1.73 m 2 , measured twice, 90 days apart) receiving care in 130 facilities across the Veterans Affairs Health Care System were included (n=281,223). We studied the proportions of patients (facility-level) receiving recommended core measures and facility-level variations of these study outcomes using median rate ratios, adjusting for various patient and provider-level factors. Median rate ratio quantifies the degree to which care may vary for similar patients receiving care at two randomly chosen facilities, with ,1 being no variation and .1.2 as substantial variation between the facilities. Study outcomes included measurement of urine albumin-to-creatinine ratio/ urine protein-to-creatinine ratio and blood hemoglobin concentration, prescription of statins and angiotensinconverting enzyme inhibitors/angiotensin receptor blockers, BP,140/90 mm Hg, and referral to a Veterans Affairs nephrologist (only for those with eGFR,30 ml/min per 1.73 m 2). Results Among those with eGFR 30-59 ml/min per 1.73 m 2 , proportion of patients receiving recommended core measures (median and interquartile range across facilities) were 37% (22%-47%) for urine albumin-to-creatinine ratio/urine protein-to-creatinine ratio, 74% (72%-79%) for hemoglobin measurement, 66% (62%-69%) for angiotensin-converting enzyme inhibitor/angiotensin receptor blocker prescription, 85% (74%-87%) for statin prescription, 47% (42%-53%) for achieving BP,140/90 mm Hg, and 13% (7%-16%) for meeting all outcome measures. Adjusted median rate ratios (95% confidence intervals) were 5.2 (4.1 to 6.4), 2.4 (2.1 to 2.6), 1.3 (1.2 to 1.3), 1.2 (1.2 to 1.3), 1.4 (1.3 to 1.4), and 4.1 (3.3 to 5.0), respectively. Median rate ratios were qualitatively similar in an analysis restricted to those with eGFR 15-29 ml/min per 1.73 m 2. Conclusions Among patients with diabetes and CKD, at facility-level, ordering of laboratory tests, and scheduling of nephrology referrals in eligible patients remains suboptimal, with substantial variations across facilities.