Screening to predict serious diseases in the general population has been regarded as unethical as it is supposed to make people anxious. Therefore we have evaluated whether mothers become anxious when their babies participate in a project to predict diabetes in the general child population. Out of 21 700 newborn children, 16 300 (75%) entered the ABIS project (All Babies in South-east Sweden). The parents (usually the mothers) answered a questionnaire at the child's birth and then again after 1 yr. A total of 10 868 representative birth questionnaires had been analyzed. To the question, 'How do you feel when you know that your child is participating in this study?', only 2.5% of mothers of children with type 1 diabetes in the family answered 'more anxious/much more anxious', and even fewer (1.5%) of the mothers in the general population (p < 0.01). A total of 52.5% of the general population answered 'calmer/more reassured' (29.3% 'calmer' and 23.2% 'much calmer'), while 43.3% felt unaffected. Those 1.5% of mothers who reported becoming more anxious were more likely to be in an unstable social situation (unemployed, p < 0.001; born abroad, p < 0.001; low education, p < 0.001). At the 1-yr follow-up, 4948 unselected questionnaires had been analyzed. Only 1.2% of the mothers felt 'more anxious', while the overwhelming majority felt either unaffected (58.7%) or calmer (38.6%). At this follow-up, most of those who had felt 'more anxious' at birth did not feel that way any longer, and none of those with diabetes in the family. We conclude that large-scale screening studies for the prediction of diabetes in the general population can be performed without causing increased anxiety. A few parents, most often found in the group with known social problems, might need extra information and support.