Starch in paper coatings is known to increase the risk of print mottle in lithographic offset printing. The objective of this study was to increase the understanding of this behaviour. Four phenomena that could lead to print mottle, where the presence of starch might be important, have been identified: uneven binder migration, uneven coating mass distribution, uneven deformation during calendering and differential shrinkage. The latter three were investigated in this project. Starch-containing coating colours often have high water retention. A relationship between the water retention of the coating colours and the distribution of coating thickness was found in a pilot trial. A theory is proposed, where the surface profile of the base paper beneath the blade, that governs the coat weight distribution in blade coating, is affected by moisture from the dewatering coating colours and the compressive force exerted by the blade. Drying strategies were studied to see whether they would induce porosity variations in the coating layers. There is a strong connection between the rate of evaporation and the shrinkage of the coating layer, but no porosity variations due to the choice of drying strategy were found. Shrinkage is governed by the capillary forces. At the same capillary pressure, the coating shrinks more for some binder systems, which is suggested to be due to a weaker chemical interaction between the binder and the pigment. Oxidized starch/latex coatings, stained with a fluorescent marker, had a greater standard deviation in fluorescence intensity than CMC/latex coatings caused by a difference in either porosity or latex distribution. It was shown that calendering introduces porosity variations into the coating layer that are larger for starch-containing coatings. The drying strategies appeared to have a significant effect on these porosity variations and they correlated positively with print mottle in some cases and in another case negatively. In the case of the negatively correlated, the mottle was probably caused by variations in surface porosity existing prior to the calendering.