The objective of our study was to retrospectively review one institution's cases of nephrogenic systemic fibrosis (NSF), evaluate possible associated factors, determine the prevalence of NSF, and search for gadolinium in skin samples obtained from patients with NSF.A retrospective review of our dermatopathology database from 1997 to 2007 was performed to search for patients with NSF. The records of patients with NSF were reviewed for factors suspected to be associated with NSF such as acidosis, low hemoglobin levels, low serum calcium levels, inflammatory conditions, serum antibodies, pharmaceutical erythropoietin, angiotensin-converting enzyme inhibitors, gadolinium-based contrast agents (GBCAs), renal failure, and dialysis. The biopsy samples from NSF patients and from control subjects were examined with energy-dispersive X-ray spectroscopy to detect gadolinium. Retrospective chart reviews of patients evaluated at our local dialysis center and our dermatology clinic were conducted to identify patients who underwent MRI, who had NSF managed exclusively by our tertiary referral centers, or both from 1997 to 2007.Seven cases of NSF were found in the dermatopathology database. Two of the seven patients were also followed up at our outpatient dialysis clinic. No other cases of NSF were discovered within the dialysis clinic's population exclusively followed within our institution. All seven dermatopathology database NSF patients developed symptoms of NSF after receiving GBCAs during renal failure and showed concomitant proinflammatory conditions. No other proposed risk factors were uniformly present in these NSF cases. All four NSF patients with chronic renal failure developed NSF after hemodialysis, with one patient dialyzed 12 hours after receiving a contrast dose. Gadodiamide was the only GBCA that all seven NSF patients received before symptom onset. Symptom onset was from 3 weeks to 18 months after GBCA exposure, with cumulative GBCA doses ranging from 0.16 to 0.43 mmol/kg. Gadolinium was detected in six of seven NSF patients' skin biopsies. Seven of eight random control specimens obtained from three healthy control subjects, three patients with renal insufficiency who had not been exposed to gadodiamide, and two patients without renal disease who had been exposed to gadodiamide were negative. Seventy-two dialysis clinic patients underwent 127 contrast-enhanced MR examinations from 1997 to 2007. Eighteen patients received gadopentetate, none of whom developed NSF. Sixty-three patients received gadodiamide, two of whom developed NSF (prevalence of NSF in patients exposed to GBCA, 2.8%; odds ratio, 0.82 [95% CI, 0.04-18.10]; likelihood ratio, 1.16 [95% CI, 1.06-1.26]). Nine patients received both contrast agents.An association with GBCAs in the development of NSF is suggested in the setting of renal insufficiency, but other factors seem to play a role. Dialysis did not prevent the development of NSF. Gadolinium was detected in skin samples from NSF patients.