BACKGROUND: Based on an inception cohort of 103 patients who had local recurrence (Group I) and a witness group of 311 patients who achieved local control (Group II) after vertical partial laryngectomy for Stage I-II glottic carcinoma, the current retrospective study documented the consequences and management of local recurrence. METHODS: Three hundred two patients (97.1%) in Group II and all 103 patients (100%) in Group I were followed until death or for a minimum of 10 years. Statistical analysis of survival, lymph node control, and distant metastasis was based on the Kaplan-Meier product limit method. RESULTS: The 10-year actuarial survival estimate was 30.8% for Group I patients and 63.1% for Group II patients. Survival was statistically more likely to be reduced in Group I patients (P < 0.0001) than in Group II patients. The percentage of patients who died of their initial disease was 44.6% in Group I and 6.3% in Group II. The 10-year actuarial lymph node control estimate was 70.2% for Group I and 96.1% for Group II. Lymph node recurrence was statistically more likely to occur in Group I patients than in Group II patients (P < 0.0001). The 10-year actuarial estimate for patients without distant metastasis was 80.2% for Group I and 96.7% for Group II. Distant metastasis was statistically more likely to occur in Group I patients than in Group II patients (P < 0.0001). Salvage treatment was unsuitable for 4.7% of patients with local recurrence; for other patients, it yielded a 86.7% local control rate, a 21.4% laryngeal preservation rate, a 4.5% death rate, and an 11.2% rate of incidence of severe complications. CONCLUSIONS: Among patients with Stage I-II glottic carcinoma managed with vertical partial laryngectomy, local recurrence results in a reduced rate of survival as well as a high rate of necessity for salvage total laryngectomy.