BACKGROUND: Inverse associations of birth length with suicide attempts have recently been reported. Whether growth during childhood alters this association is not known. The influences on patterns of growth in fetal life and childhood might be different for violent and non-violent suicide attempts. OBJECTIVE: To investigate the effect of fetal and childhood growth and possible effect modification on suicide attempts, both violent and non-violent, adjusting for potential maternal confounding factors. METHOD: 318 953 Men were followed by record linkage from the date of birth in Sweden (1973-1980) to the date of attempted suicide, date of death, emigration or to the end of 1999. RESULTS: The risk of suicide attempt was increased for men with reduced linear growth in fetal life across all levels of adult stature. Men with appropriate birth length for gestational age but short adult height also experienced a raised risk of suicide attempts: 1.56 (95% CI 1.2 to 2.1). Tall adult stature was protective. Short birth length for gestational age was more strongly related to violent (2.39; 95% CI 1.1 to 4.9) than non-violent (1.53; 95% CI 1.1 to 2.1) suicide attempts. The risk of violent attempts was most strongly increased for men with low birth weight and adequate adult stature: 2.54 (95% CI 1.1 to 5.7). CONCLUSIONS: The inverse association of linear growth in fetal life and suicide attempt does not seem to be modified by linear childhood growth. Short adult stature entails an additional risk. Short birth length seems particularly to increase the risk of violent suicide attempts.