The data gathered in connection with a routine health screening program for 4-year-olds were used in order to test a number of hypotheses concerning factors which might have influenced the effectiveness of the program, as well as concerning the quality of preventive care delivered to the children before the age of 4 years. It was found that preventive care of relatively high quality delivered to children 0-3 years of age had not reduced the prevalence of previously undetected health problems at 4 years of age. Furthermore, the analysis demonstrated substantial differences between physicians in the rates of correct and unnecessary referrals. Only a small part of these differences could be attributed to professional status or specific experience with the program. They consisted essentially in varying thresholds for what was perceived as a health problem needing treatment, and were considered as a matter of personal temperament. True differences of quality did occur but were less prominent. The implications of these findings with regard to the continuation of the program are discussed.