We randomized patients with chronic kidney disease (serum creatinine ≥ 1.5 mg/dl or glomerular filtration rate (GFR) <60 ml/min/1.73 m²) in a double-blind fashion to receive saline or sodium bicarbonate prior to and after cardiac or vascular angiography. The primary endpoint was contrast-induced nephropathy (CIN), defined as an increase in serum creatinine by 25% or by 0.5 mg/dl from baseline. Patients with congestive heart failure (CHF), cardiac ejection fraction (EF) <30%, or GFR < 20 ml/min/1.73 m² were excluded. The study was discontinued (after 142 patients were randomized) due to a low incidence of CIN (1.5%). We retrospectively identified all cases of CIN (n = 30) at our institution during the same time period to see if these patients differed from our trial sample. There was no difference in serum creatinine (1.7 ± 0.4 vs. 1.7 ± 0.6 mg/dL), GFR (42.7 ± 9.7 vs. 45.3 ± 3.2 ml/min), incidence of diabetes (51.8% vs. 63.3%), contrast volume (121.7 ± 63.8 vs. 122.7 ± 68.3 ml), ACE inhibitor or angiotensin receptor blocker use (54.0% vs 63.3%), and periprocedure diuretic use (33.1% vs 26.7%). On multivariate analysis, only a cardiac ejection fraction (EF) of less than 40% was significantly associated with CIN (odds ratio, 4.52; 95% confidence interval, 1.30-15.71; P = 0.02). In all, 22/30 patients (73.3%) who developed CIN had at least one or more characteristics that would have excluded their enrollment in our randomized trial including evidence of congestive heart failure (17/30 patients), EF less than 30% (9 patients), age greater than 85 years (2 patients), or advanced renal failure with a baseline GFR of less than 20 cc/min (1 patient). In summary, patients with CKD without evidence of CHF who receive adequate hydration appear to have a very low risk of CIN associated with angiography. A low EF (less than 40%) appeared to be the most significant risk factor for CIN in our population.