AIMS--To determine whether the abundance of newly formed collagen in healing surgical wounds correlated with scar severity, and whether topical application of steroid cream reduced new collagen formation in patients who have undergone median sternotomy. METHODS--Thirty three patients six weeks after sternotomy, and 12 controls were studied. Scars were photographed, and biopsy specimens from scars at sites treated or untreated with topical corticosteroids (clobetasol proprionate 0.5%) were examined using immunohistochemical staining for type 1 procollagen (PCP 1) and transforming growth factor beta (TGF-beta), and in situ hybridisation for type 1 procollagen messenger RNA (mRNA). RESULTS--The degree of hypertrophy of the scar and the abundance of PCP 1 immunostaining were ranked independently, blind, and a correlation between these two variables was observed (r = 0.604, p < 0.001). The PCP 1 immunostaining was accompanied by a great abundance of PCP 1 mRNA and only a slight increase in TGF-beta immunostaining, when compared with normal skin or mature scars. Following the application of topical corticosteroids, for either 48 hours or twice daily for seven days, there was no reduction in PCP 1 immunostaining nor the abundance of PCP 1 mRNA. CONCLUSIONS--These data suggest that the extent of new collagen formation as assessed by PCP 1 immunohistochemistry may be a useful marker of the exuberance of the scarring process following sternotomy, and that topical corticosteroids are ineffective in reducing this component of the fibrotic response.