We reviewed the medical records of 98 children with Crohn's disease followed at Children's Hospital of Pittsburgh from 1983 to 1993 to evaluate the merits of alternate-day prednisone (AD) maintenance therapy once initial remission was achieved. Of the 98 children, 35 had adequate data recorded for eligibility to the study. Of these, 11 were in the AD group and 24 were in a group whose maintenance regimen did not include prednisone (NO). The dependent variables were frequency of flares and linear growth over time. AD therapy reduced mean symptomatic flares (0.23 +/- 0.1 vs 0.69 +/- 0.14 flares/patient/year; p = 0.04) over a 2-year follow-up period but did not delay significantly the onset of a flare after remission was achieved (16.5 +/- 3.4, vs 13.4 +/- 1.8 months; p = 0.4). Site of disease involvement had no impact on frequency of flares. Fewer patients in the AD group experienced flares, but this finding did not achieve statistical significance (4/11, 36%, vs 17/24, 71%; p = 0.07). Linear growth, measured in height percentile and growth velocity (cm/year), was not significantly reduced by the second year of either therapy. This small retrospective study suggests that AD prednisone therapy may be effective in reducing symptomatic flares in Crohn's patients without a resultant inhibition of linear growth.