A growing number of dermatologists dispute the existence of infantile seborrheic dermatitis (ISD) as an independent clinical entity. Therefore the aim of the present study was to estimate the epidemiologic features of ISD in a defined population of Greek children, assess its course, and identify associations, if any, with other common dermatoses of childhood. Children from the region of Athens who were examined and diagnosed with typical clinical features of ISD between 1997 and 2011 were included in the study. The relevant data were collected retrospectively from their medical records using a standardized form. Eighty-seven children were enrolled (50 boys, 37 girls; mean age 3.1 mos at the time of ISD diagnosis). The main body areas affected were the scalp and face for the majority of the children (78/87), whereas the trunk and limbs were less frequently involved (9/87). In all cases, erythema and scaling of affected patients were mild to moderate. Forty-nine of the 87 children were followed up over a period of 5 years. Thirty children in this group developed features of atopic dermatitis (AD) at a later stage, according to the UK diagnostic criteria of AD, and 23 of these children were diagnosed with AD, at an average time interval of 6.4 months from ISD onset, and seven presented with clinical features of AD at the time of ISD diagnosis. The remaining 19 children in the follow-up group progressed without developing any other chronic skin disease, and all recovered within 6 months of its onset. Thirty-eight had no further follow-up after their initial ISD diagnosis. In spite of the lack of information on the disease course for the last group, assuming they all recovered, the prevalence of AD (34.4%) in our ISD sample was significantly higher than the prevalence of AD (10.7%) in the general population for the same age group, as shown in a previous study performed in the municipality of Athens (p < 0.001). A significant number of children were found to develop AD shortly after their ISD diagnosis. This finding demonstrates a strong association in the clinical course between the two diseases or indicates that the two diseases may be in the same clinical spectrum. Further epidemiologic studies must be conducted to assess the significance of this finding.