Alternate-day steroids are currently recommended to treat children with membranoproliferative glomerulonephritis (MPGN). This recommendation is based largely on uncontrolled studies demonstrating improved renal survival with steroid therapy. We reviewed the outcome of 39 children who presented with MPGN between 1968 and 1990; 27 children were treated with steroids and 12 children received no drug therapy. Life-table analysis comparing renal survival of treated versus untreated children demonstrated no difference by log rank analysis. Treated and untreated groups were compared on the basis of nine features at presentation: age, sex, type of MPGN, presence and type of hematuria, hypocomplementemia, renal insufficiency, hypertension, and nephrosis. Treated children were likely to be female (P<0.01) and nephrotic (P<0.02). Actuarial survival analyses were performed comparing the nine features with renal survival through 10 years of follow-up. Normotensive (P<0.025) and non-nephrotic (P<0.05) children had improved renal survival. The 11 non-nephrotic children demonstrated 100% long-term renal survival, including 7 who received no steroid therapy. At last follow-up, all non-nephrotic children had normal renal function, serum albumin levels >3 g/dl, and were normotensive. These data suggest that non-nephrotic children with MPGN may forego steroid treatment without compromising long-term renal function. The current common practice of treating all children with MPGN with steroids should be re-examined. © 1995 IPNA.