Objectives Methyl chloride leakage from a refrigerator occurred on board an Icelandic fishing vessel in the year 1963. The exposure lasted four days and one man died during the fishing trip. Many of the crew members were hospitalised due to different neurological symptoms and signs, and had not recovered completely several years later. The aims were to study long-term mortality.

Methods This is a cohort study with external references. Five referents were randomly selected from registries of seamen and officers matched to each crew member according to age and employment status. Follow up was through record linkage of personal identifier with nationwide mortality registry. Hazard ratios (HR) and 95% confidence intervals (CI) were estimated in Cox proportional hazards model adjusted for age and employment.

Result The intoxicated crew eligible for follow up were 20 deckhands and 7 officers, the reference group counted 100 deckhands and 35 officers. Followed up to end of 2010, 14 of the exposed deckhands had died and 6 of the officers versus 49 deckhands and 26 officers among the reference group. The HR for all causes of death was 2.10 (95% CI 1.28–3.46). For all cardiovascular events HR was 2.06 (95% CI 1.02–4.15), for acute coronary heart disease HR was 3.12 (95% CI 1.11–8.78), for cerebrovascular diseases HR was 5.35 (95% CI 1.18–24.35), and for suicide HR was 13.76 (95% CI 1.18–160.07).

Conclusions The follow up of the methyl chloride exposed cohort showed increased mortality due to cardiovascular diseases after 47 years. After the intoxication the suicide cases had developed severe depressions that were considered to be related to the exposure. The use of the personal identifiers and the comprehensive population registries strengthen the study. Comparison to non exposed group of the same occupations indirectly control for potential confounders including social class, occupational experience, lifestyle factors, diet, smoking, and alcohol use.