To study the evolution of ochratoxin A (OTA) content from must to wine during the making of Port Wine, grapes from the five most common varieties of Port Wine were harvested and combined in equal percentages in order to perform microvinifications. Three sets of assays were studied: a blank (A), where the most common Port Wine-making process was used; in the second (B), a solution of OTA was added to the initial must; in the third (C), the grapes were aspersed with an inoculating solution of OTA-producing fungi. Samples were collected, in duplicate, on four different occasions throughout the process. The influence of the addition of SO2 to the must was also assessed in each set. The quantification of OTA was based on the standard reference method for wines (European Standard prEN 14133), which includes clean-up via immunoaffinity columns and HPLC with fluorescence detection. The limits of detection were 0.076 microg/l for wine and 0.114 microg/l for must. The method was validated by assessing the precision, accuracy and by obtaining an estimate of the global uncertainty. Overall, the levels of OTA observed during the vinifications dropped by up to 92%, and no grapes used in this work were contaminated naturally.