Inflated-fixed lung specimens, prepared using polyethyleneglycol 400, the standard Heitzman method have the disadvantage of bad staining. The authors attempted to improve this short-coming using human autopsied lungs and evaluated both microscopical findings and the degree of specimen inflation. Fifty-five human autopsied whole lungs were examined. Forty-six lungs with cannulation through the main bronchus were distended with 20% formalin, after clamping the cannula for 15 minutes to six days, formalin in lungs was expelled by air at 10 cm H2O pressure, then fixed with fixative-fluid containing polyethyleneglycol 400, 95% ethyl-alcohol, 40% formalin and water mixed in a ratio of 10:5:2:3. In 9 lungs without cannulation, formalin was expelled manually. The degree of distension of 46 of the 55 cannulated specimens was satisfactory (score A: 42 specimens, score B: 4 specimens). In 9 specimens without cannulation, inflation was poor (Score C). Good or excellent staining was obtained in 29 specimens (Score A: 10 specimens, Score B: 19 specimens), but 3 specimens were not be improved (Score C). Both inflation and staining were good in 21 specimens. This new method, requiring cannulation in the main bronchus, inflated-fixation with 20% formalin and expulsion of formalin by air prior to the standard Heitzman method enables both high quality staining and radiographs of the inflated-fixed lung specimen.