Objective-To assess physiological cardiac adaptation in adolescent professional soccer players. Subjects and design-Over a 32 month period 172 teenage soccer players were screened by echocardiography and ECG at a tertiary referral cardiothoracic centre. They were from six professional soccer teams in the north west of England, competing in the English Football League. One was excluded because of an atrial septal defect. The median age of the 171 players assessed was 16.7 years (5th to 95th centile range: 14-19) and median body surface area 1.68 m 2 (1.39-2.06 m 2 ). Main outcome measures-Standard echocardiographic measurements were compared with predicted mean, lower, and upper limits in a cohort of normal controls after matching for age and surface area. Univariate regression analysis was used to assess the correlation between echocardiographic variables and the age and surface area of the soccer player cohort. ECG findings were also assessed. Results-All mean echocardiographic variables were greater than predicted for age and surface area matched controls (p < 0.001). All variables except left ventricular septal and posterior wall thickness showed a modest linear correlation with surface area (r = 0.2 to 0.4, p < 0.001); however, left ventricular mass was the only variable that was significantly correlated with age (r = 0.2, p < 0.01). Only six players (3.5%) had structural anomalies, none of which required further evaluation. All had normal left ventricular systolic function. Sinus bradycardia was found in 65 (39%). The Solokow-Lyon voltage criteria for left ventricular hypertrophy were present in 85 (50%) and the Romhilt-Estes points score (five or more) in 29 (17%). Repolarisation changes were present in 19 (11%), mainly in the inferior leads. Conclusions-Chamber dimensions, left ventricular wall thickness and mass, and aortic root size were all greater than predicted for controls after matching for age and surface area. Sinus bradycardia and the ECG criteria for left ventricular hypertrophy were common but there was poor correlation with echocardiographic left ventricular hypertrophy. The type of hypertrophy found reflected the combined endurance and strength based training undertaken. (Heart 2001;85:649-654)