To determine whether the ductal carcinoma in situ (DCIS) detected mammographically or presenting clinically is the same or differs, pathological and biological (c-erbB-2 and p53 detection) features of 79 cases of pure DCIS, 5 cases with microinvasion and 8 cases with 1 to 2 mm of invasion, all detected by mammography, have been compared with 59 cases of pure DCIS, 8 cases with microinvasion and 7 cases with 1 to 2 mm invasion, all of which presented clinically. Half of the mammographically detected group were smaller than 20 mm, and there was a higher incidence of these being low grade, whereas 30% of the symptomatic cases were smaller than 20 mm, and more of this group were larger than 50 mm. For the pure DCIS, there were less high-grade and more intermediate-grade cases in the mammographically detected group, although the incidence of low grade was similar between the two groups. There were more cases with a micropapillary pattern in the symptomatic group. C-erbB-2 protein was detected in 42% of the mammographically detected cases, whereas 59% of the symptomatic cases had c-erbB-2 reactivity. P53 detection was similar for both groups (33.0% and 37.0%). There were more symptomatic cases with invasion, and these were predominantly high grade, whereas the mammographically detected cases were both high and intermediate grade. Twelve of the 15 symptomatic cases with invasion expressed c-erbB-2, in comparison with 4 of the 13 mammographically detected cases, with half of the high-grade lesions in the latter group being negative. This study has shown that although there is overlap of pathological and biological features between DCIS presenting clinically and that detected mammographically, there can be differences in extent, grade, and invasion. The impact of this, however, can be determined only by clinical follow-up.