From the early 1970's to the late 1980's, Pacific Northwest National Laboratory conducted life-span studies in beagle dogs on the biological effects of inhaled plutonium ((238)PuO(2), (239)PuO(2), and Pu[NO(3)](4)) to help predict risks associated with accidental intakes in workers. Years later, the purpose of the present follow-up study was to reassess the dose-response relationship for lung cancer in the PuO(2) dogs compared to controls-with particular focus on the dose-response at relatively low lung doses. A PuO(2) aerosol (2.3 mum activity-median aerodynamic diameter, 1.9 mum geometric standard deviation) was administered to six groups of 20 young (18-mo-old) beagle dogs (10 males and 10 females) by inhalation at six different activity levels, as previously described in Laboratory reports. Control dogs were sham-exposed. In dose level 1, initial pulmonary lung depositions were 130 + or - 48 Bq (3.5 + or - 1.3 nCi), corresponding to 1 Bq g lung tissue (0.029 + or - 0.001 nCi g(-1)). Groups 2 through 6 received initial lung depositions (mean values) of 760, 2,724, 10,345, 37,900, and 200,000 Bq (22, 79, 300, 1,100, and 5,800 nCi) PuO(2), respectively. For each dog, the absorbed dose to lungs was calculated from the initial lung burden and the final lung burden at time of death and lung mass, assuming a single, long-term retention function. Insoluble plutonium oxide exhibited long retention times in the lungs. Increased dose-dependent mortality due to lung cancer (bronchiolar-alveolar carcinoma, adenocarcinoma, and epidermoid carcinoma) and radiation pneumonitis (in the highest exposure group) were observed in dogs exposed to PuO(2). Calculated lung doses ranged from a few cGy (lowest exposure level) to 7,764 cGy in dogs that experienced early deaths from radiation pneumonitis. Data were regrouped by lifetime lung dose and plotted as a function of lung tumor incidence. The lung tumor incidence in controls and zero-dose exposed dogs was 18% (5/28). However, no lung tumors were observed in 16 dogs with the lowest lung doses (8 to 22 cGy, mean 14.4 + or - 7.6 cGy), and only one lung tumor was observed in the next 10 dogs with lung doses ranging from 27 to 48 cGy (mean 37.5 + or - 10.9 cGy). By least-squares analysis, a pure-quadratic function represented the overall dose-response (n = 137, r = 0.96) with no apparent dose-related threshold. Reducing this function to three linear dose-response components, we calculated risk coefficients for each. However, the incidence of lung tumors at zero dose was significantly greater than the incidence at low dose (at the p < or = 0.053 confidence level), suggesting a protective effect (radiation homeostasis) of alpha-particle radiation from PuO(2). If a threshold for lung cancer incidence exists, it will be observed in the range 15 to 40 cGy.