Resected non-small cell lung cancer (NSCLC) has 5-years survival rates of 30-70%. The role of adjuvant chemotherapy remains unclear with poor compliance reported in most trials. The compliance with adjuvant chemotherapy (ACT) for stage IB and II NSCLC was analyzed using data from a North American multi-centre phase III study (accrual 1994-2001) that compared adjuvant chemotherapy to observation. Planned chemotherapy consisted of cisplatin (CIS) 50 mg/m2 days 1, 8 and vinorelbine (VIN) 25 mg/m2 days 1, 8, 15, 22 for four cycles; the VIN dose had been reduced from 30 mg/m2 after an initial cohort of patients experienced unacceptable toxicity. Four hundred and twenty-four patients were randomized after the amendment, 215 to the chemotherapy arm. Median age was 60 years, 64% were male and 84% had stage II disease. Thirty-seven patients completed one cycle, 14 completed two, 20 completed three and 108 patients completed all four cycles. Ten patients received no therapy. Multivariate analysis demonstrated statistically significant differences in compliance with extent of surgery, gender and age. Patients randomized in Canada were more likely to fail to complete chemotherapy due to refusal of therapy than their American counterparts. Patients who had pneumonectomies were more likely to discontinue therapy due to toxicity than those who had lesser resections. Extent of surgery may play a role in both the compliance and toxicity of ACT. Differences between nations in the perception of the risks and benefits of adjuvant chemotherapy regimens, both between physicians and patients, should be investigated further.