The Vietnam Era Twin (VET) Registry includes 14 800 male twins born 1939–55 and in military service in 1964–75. A mailed health survey including the Jenkins Sleep Questionnaire was sent to 11 959 members and 8870 (74.2%) provided responses on the frequency of sleep problems in the previous month. Prevalence of those experiencing conditions at least 1 day per month was 67.2% for waking often, 61.5% for waking tired/worn out, 48.1% for trouble falling asleep and 48.6% for awakening early. Ordinal logistic regression analysis was used to estimate sleep problems from demographic, behaviour and lifestyle characteristics, and morbid conditions. Black (vs. white) race, older age, church/religious group participation, social support, employment, cigarette smoking, light physical activity, and strenuous physical activity were associated with lower risk of one or more sleep problems. Eleven morbid conditions with a prevalence of 1% or more, coffee consumption, heavy alcohol consumption, and Framingham Type A behaviour pattern were associated with a higher risk of sleep problems. These analyses suggest that sleep problems may be one of the mechanisms relating reduced quality of life to many physical and behavioural characteristics. Fortunately, a number of the risk factors associated with sleep problems are lifestyle characteristics which, if modified, may reduce sleep problems.