This article summarises the experience of one academic rheumatologist in treatment of patients with rheumatoid arthritis (RA) over 25 years from 1980-2004 with low-dose prednisone, most with <5 mg/day over long periods. A database was available which included medications and multidimensional health assessment questionnaire (MDHAQ) scores for physical function, pain, and routine assessment of patient index data (RAPID3), completed by all patients at all visits in the infrastructure of care. Most patients were treated with long-term low-dose prednisone, often from the initial visit and indefinitely, and with methotrexate after 1990. The mean initial prednisone dose declined from 10.3 mg/day in 1980-1984 to 3.6 mg/day in 2000-2004. Although no formal criteria were used to determine the initial dose, prednisone doses were higher in patients who had more severe MDHAQ/RAPID3 scores, as expected, reflecting confounding by indication. Similar improvements were seen in clinical status over 12 months in patients treated with <5 vs. ≥ 5 mg/day prednisone, and maintained for >8 years. Adverse effects were primarily bruising and skin-thinning, with low levels of hypertension, diabetes, and cataracts, although this information was based only on self-report rather than systematic assessment by a health professional. These data reflect limitations of observational data. However, a consecutive patient database may provide long-term information not available from clinical trials. The data document that prednisone at doses <5 mg/day over long periods appears acceptable and effective for many patients with RA at this time. Further clinical trials and long-term observational studies are needed to develop optimal treatment strategies for patients with RA with low-dose prednisone.