Accurate quantification of transverse relaxation times in vivo is of vital importance for research and clinical applications. At higher field-strength, the gain in signal enables T2 mapping at sub-millimetre resolutions, but with infeasible scan time for standard spin-echo techniques. Using CPMG echo trains reduces the acquisition time. However, inhomogeneities of the transmit B1 field hamper accurate T2 quantification. Correcting for resulting bias effects is possible through signal response simulations via the Bloch equations using the specific sequence parameters. Matching acquired data to the simulated signal points allows accurate and robust fitting of T2 values as shown by our 7T study.