The objective of our study was to relate alterations in biceps tendon diameter and signal on MR images to gross anatomy and histology. T1-weighted, T2-weighted fat-saturated, and proton density-weighted fat-saturated spin-echo sequences were acquired in 15 cadaveric shoulders. Biceps tendon diameter (normal, flattened, thickened, and partially or completely torn) and signal intensity (compared with bone, fat, muscle, and joint fluid) were graded by two readers independently and in a blinded fashion. The distance of tendon abnormalities from the attachment at the glenoid were noted in millimeters. MRI findings were related to gross anatomic and histologic findings. On the basis of gross anatomy, there were six normal, five flattened, two thickened, and two partially torn tendons. Reader 1 graded nine diameter changes correctly, missed two, and incorrectly graded four. The corresponding values for reader 2 were seven, one, and five, respectively, with kappa = 0.75. Histology showed mucoid degeneration (n = 13), lipoid degeneration (n = 7), and fatty infiltration (n = 6). At least one type of abnormality was found in each single tendon. Mucoid degeneration was hyperintense compared with fatty infiltration on T2-weighted fat-saturated images and hyperintense compared with magic-angle artifacts on proton density-weighted fat-saturated images. MRI-based localization of degeneration agreed well with histologic findings. Diameter changes are specific but not sensitive in diagnosing tendinopathy of the biceps tendon. Increased tendon signal is most typical for mucoid degeneration but should be used with care as a sign of tendon degeneration.