T2 measurements of 31P NMR signals of ATP using the Hahn 90 degrees-180 degrees spin-echo sequence imply difficulties whenever the 180 degrees pulse angle is not completely perfect. The reason for this finding is the crucial influence of the J-couplings of the ATP signals which result in intensity modulations and consequently in false T2 values even when the echo times are chosen to TE = n/J. Examinations on the calf muscles of healthy volunteers were performed to demonstrate this effect and its influence on in vivo T2 determinations. The T2 relaxation times evaluated with the Hahn spin-echo in combination with a Helmholtz coil are far shorter than the true T2 values.