The aim of this study was to investigate the influence of the shelling process on the presence of ochratoxin A (OTA) in cocoa samples. Twenty-two cocoa samples were analysed for the determination of OTA before (cocoa bean) and after undergoing manual shelling process (cocoa nib). In order to determine OTA contamination in cocoa samples, a validated high-performance liquid chromatography (HPLC) method with fluorescence detection was used for the quantitative analysis of ochratoxin A (OTA). In both types of samples, OTA was extracted with methanol-3% sodium hydrogen carbonate solution and then purified using immunoaffinity columns prior to HPLC analysis. Due to the fact that different recovery values were obtained for OTA from both types of samples, a revalidation of the method in the case of cocoa nibs was needed. Revalidation was based on the following criteria: Selectivity, limits of detection and quantification (0.03 and 0.1 microg kg(-1), respectively), precision (within-day and between-day variability) and recovery 84.2% (RSD = 7.1%), and uncertainty (30%). Fourteen of the twenty-two cocoa bean samples (64%) suffered a loss of OTA of more than 95% due to shelling, six samples suffered a loss of OTA in the range 65-95%, and only one sample presented a reduction of less than 50%. The principal conclusion derived from this study is that OTA contamination in cocoa beans is concentrated in the shell; therefore, improvements of the industrial shelling process could prevent OTA occurrence in cocoa final products.