Alveolar type II cell tumors were induced transplacentally by intraperitoneal injection of pregnant C3H/HeNCr MTV- or Swiss Webster mice with N-nitrosoethylurea at a dose of 0.5 mmol/kg and 0.74 mmol/kg. At different time points after birth (1-32 weeks), the entire lungs from 40 of the male offspring were inflated with Bouin's fixative, separated into lobes, and sectioned at 5 microns serially to detect every microscopic lesion. Results were compared with those obtained from examining only every 10th, 20th, or a single midlevel section from the same material. On average, 150 serial sections were prepared per mouse lung. Initially, only purely solid/alveolar or purely tubulopapillary types were observed but with tumor progression, papillary structures developed within solid tumors resulting in mixed neoplasms. Analyzing mouse lungs in step sections of every 10th section (50-60 microns), 5/238 (2%) of the tumors were missed, in step sections of every 20th section (100-120 microns), 16/238 (7%) of the tumors were not detected and usually less than half of the tumors were seen in the single mid-level section. The approximate size of the neoplasms is indicated by the total number of sections per tumor. The dimensions of tumors evaluated with step sections of 10 or 20 were comparable to the size observed with serial sections. It is concluded that the evaluation of mouse lung tumors in steps of approximately 50 microns is basically equivalent to the study of serial sections and appears to be a feasible method to assess the complete incidence, histological type, and size of all proliferative processes throughout the entire lung.