This article reviews the design and findings of LCSG Protocol 771, a randomized double-blind comparison of postoperative adjuvant intrapleural bacillus Calmette-Guérin (BCG) against saline solution placebo control in 473 patients with resected T1N0, T1N1, or T2N0 non-small cell lung cancer. Patients were randomized from August 30, 1977, through October 20, 1980, and follow-up ended on January 1, 1990. There was no evidence of improved survival or time to recurrence among patients given BCG in contrast to earlier promising findings. A calculation suggests that false-positive results due to chance are not uncommon in small preliminary studies, indicating the need for larger confirmatory trials such as LCSG Protocol 771. Other contributions from this Protocol are also reviewed, including data on BCG toxicity, immunologic effects, patterns of recurrence, and identification of prognostic risk groups.