The purpose of this study was to determine whether there are any relationships between the morphological expressions of the progression of neoplasia in tracheal epithelium in vivo, and the increased in vitro growth autonomy expressed by carcinogen-altered cell populations isolated from the same tracheas. Rat tracheal implants were exposed for 2 weeks or 4 weeks to 200 micrograms 7,12-dimethylbenz[a]anthracene (DMBA)-beeswax pellets. In the first phase of this study reported earlier (1), the numbers of carcinogen-altered cell populations, identified by their lack of need for exogenous pyruvate and insulin for survival in cell culture, were quantitated 2, 6 and 9 months after the start of the exposures. Before generating the cell cultures, lesions on the pieces of pre-exposed tracheal implants were identified by placing them in organ culture for 24 h and collecting the exfoliated cells from the medium for diagnostic cytopathology. Although the pieces of 2-week DMBA-exposed tracheas had not developed any markedly atypical lesions by 9 months, there was a progressive increase in the growth autonomy of the small number of subculturable cell populations (approximately 1/trachea) obtained from this exposure group. This was seen in the decreased time needed in culture to expand the population for testing anchorage-independent growth and tumorigenicity in nude mice. Also, anchorage-independent growth was markedly enhanced from 25 to 80% between 2 and 6 months and then to 100% at 9 months. Tumorigenicity did not show an increase. In the 4-week DMBA-exposed group, 74% of the large number of subculturable cell populations (approximately 5.0/trachea) already showed anchorage-independent growth at 2 months. This percentage increased to 94-100% at 6 and 9 months. A progressive increase in tumorigenic cell populations was also clearly seen. At 2 months, 26% of the cell populations inoculated into nude mice formed tumors. This number increased from 45 to 61% at 6 and 9 months, respectively. The pieces of tracheas taken at 2 months after the start of the 4-week DMBA exposure had no markedly atypical lesions. At 6 months one marked atypia was detected, and at 9 months five markedly atypical lesions and two carcinomas in situ were present. The cell populations derived from the explants harboring these lesions showed a high incidence of growth autonomy, e.g. anchorage-independent growth and tumor formation, indicating a close relationship between these properties and the evolving morphological manifestations of tumor development.