OBJECTIVE:To compare patellar tendon sonographic findings at baseline and at follow-up in active female basketball players with and without symptoms of jumper's knee. We hypothesized that baseline sonographic morphology would not reliably predict prognosis and, in particular, that it would not predict the need for surgery. DESIGN:Prospective longitudinal study with 12-month minimum follow-up. SETTING:Institutional elite athlete study group in Australia (Victorian Institute of Sport Tendon Study Group). PATIENTS AND PARTICIPANTS:A total of 15 female elite basketball players with 23 sonographically abnormal tendons and 15 matched control basketball players with 23 sonographically normal tendons. MAIN OUTCOME MEASURES:Sonographic patellar tendon appearance and clinical assessment of symptoms of jumper's knee at baseline and follow-up. Dimensions of abnormal regions were measured. RESULTS:At baseline, the 23 subject tendons contained sonographic hypoechoic regions (six currently symptomatic, eight previously symptomatic only, and nine never symptomatic). At follow-up, the hypoechoic areas in seven tendons had resolved (and caused no symptoms), the hypoechoic areas in 11 tendons had remained essentially the same size (five were symptomatic), and the hypoechoic areas in five tendons had expanded (three symptomatic). At baseline, there were no differences between the mean +/- SD cross-sectional areas of the abnormalities in the tendons that subsequently resolved (15.9 +/- 10.1 mm2) and those that remained unchanged (39.3 +/- 25.8) or expanded (25.3 +/- 12.5). The presence of a baseline sonographic abnormality predicted symptoms of jumper's knee at follow-up (p < 0.05), but the presence of symptoms of jumper's knee at baseline also predicted symptoms at follow-up (p < 0.05). No subject or control missed any games or underwent surgical treatment. CONCLUSIONS:Patellar tendon sonographic hypoechoic areas can resolve, remain unchanged, or expand in active sports-women without predicting symptoms of jumper's knee. Thus, symptoms were not directly related to sonographic tendon morphology. Sonographic hypoechoic regions ought not to constitute per se an indication for surgery.