We report on two patients, 74 and 73 years old, each of whom had had biopsy of the bronchial mucosa elsewhere. Histologically, normal mucosa and mild bronchitis was seen, respectively. In both patients a second biopsy specimen from the same site was taken, 4 and 7 days after the first biopsy. Now the diagnosis was carcinoma in situ (not excluding early squamous carcinoma) and squamous carcinoma, respectively. One of the patients came to our lung clinic for operation on the basis of the mentioned diagnosis. A third biopsy was performed for control purposes. It revealed just a perforation of a lymph node in the bronchial wall and some epithelioid granulomas without evidence of acid-fast bacilli. There was no evidence of malignancy in this specimen. Because of the discrepancy, the pathologist at the first institution was asked to send us the previous slides. On review of these slides, the described findings were seen, but they were interpreted as regenerative-atypical. In this way, an unnecessary surgical intervention was avoided. After a follow-up of 3 1/2 years the patient is healthy. In the other patient, because of the diagnosis of squamous carcinoma a lobectomy was undertaken by a surgeon elsewhere. The pathologist could not find any carcinoma in the surgical specimen. The biopsy material was reviewed in our department, the findings being interpreted as atypical, originating in reparative changes. Like the postoperative pathologist, we did not find any carcinoma in the surgical specimen. In both cases, therefore, regenerative-atypical changes after re-biopsy were first misinterpreted as a malignant process.