A 73-year-old woman had previously been diagnosed with CREST syndrome, PBC and diabetes. Hepatic fibrosis was not evident, in spite of the transudative ascites and active esophageal varices. ACA were positive, whereas AMA and anti-gp210 antibodies were negative. She showed low urinary excretion of C-peptide and was weakly positive for anti-GAD antibody. She was diagnosed with a form of PBC that progresses via portal hypertension rather than liver failure and with SPIDDM. Her HLA type did not contain risk allele for IDDM or PBC. SPIDDM should be considered when patients with PBC with portal hypertension-type progression develop diabetes.