Internal standards have been introduced in quantitative mass spectrometry imaging in order to compensate for differences in intensities throughout an image caused by, for example, difference in ion suppression or analyte extraction efficiency. To test how well the internal standards compensate for differences in tissue types in, for example, whole-body imaging, a set of tissue homogenates of different tissue types (lung, liver, kidney, heart, and brain) from rabbit was spiked to the same concentration with the drug amitriptyline and imaged in the same experiment using isotope labeled amitriptyline as internal standard. The results showed, even after correction with internal standard, significantly lower intensities from brain and to some extent also lung tissue, differences which may be ascribed to binding of the drug to proteins or lipids as known from traditional bioanalysis. The differences, which for these results range approximately within a factor of 3 (but for other compounds in other tissues could be higher), underscore the importance of preparing the standard curve in the same matrix as the unknown sample whenever possible. In, for example, whole-body imaging where a diversity of tissue types are present, this variation across tissue types will therefore add to the overall uncertainty in quantitation. The tissue homogenates were also used in a characterization of various phenomena in quantitative MSI, such as to study how the signal depends of the thickness of the cryo-section, and to assess the accuracy of calibration by droplet deposition. For experiments on liver tissue, calibration by spiked tissue homogenates and droplet deposition was found to provide highly similar results and in both cases linearity with R2 values of 0.99. In the process, a new method was developed for preparation of standard curves of spiked tissue homogenates, based on the drilling of holes in a block of frozen liver homogenate, providing easy cryo-slicing and good quantitative performance.