Diagnosing atopic dermatitis (AD) in infants is challenging.To determine incidence and persistence of eczema and AD in infants using the UK Working Party (UKWP) and Hanifin&Rajka's (H&R) criteria.A cohort of 1834 infants were examined clinically at 3, 6 and 12 months of age. Eczema was defined as eczematous lesions excluding differential diagnosis to AD. Atopic dermatitis was diagnosed by UKWP (3, 6 and 12 months) and H&R (12 months) criteria. Logistic regression models were used to assess the relationship between AD and eczema.Eczema was observed in 628/1834 (34%) infants; 240, 359 and 329 at 3, 6 and 12 months, respectively, with AD diagnosed in 212 (34%) of infants with any eczema, and in 64/78 (82%) infants with eczema at all three visits. Odds for AD were lower with first presentation of eczema at 6 (OR: .33; 95% CI: .22 - .48) or 12 months (OR: .49; .32 - .74) compared to three months, and higher in infants with eczema at three (OR: 23.1; 95% CI: 12.3-43.6) or two (6.5; 4.3-9.9) visits compared to one visit only. At 12 months, 156/329 (47%) fulfilled the UKWP and/or H&R criteria; 27 (8%) the UKWP criteria only and 65 (20%) the H&R criteria only in whom 44/129 (34%) did not meet the itch criterion.The UKWP and H&R criteria for AD used in combination and at multiple time points may be useful in clinical research, but may have limited value in most other clinical settings.