Growing demand for fresh, unprocessed food favours the emergence of Cryptosporidium infections in humans. Mainly it is food of plant origin or unpasteurized milk which have been involved in food-borne outbreaks of cryptosporidiosis. So far consumption of shellfish contaminated with Cryptosporidium were not associated with human infections although such as possibility exists. In this study an attempt was undertaken to evaluate the analytical performance of three commonly used methods for recovery of Cryptosporidium oocysts from shellfish: i) pepsin digestion of shellfish in conjunction with immunomagnetic separation (IMS) of oocysts (method A), ii) pepsin-HCl treatment of shellfish homogenate without IMS (method B), and iii) a strainer method with direct oocyst extraction and separation from shellfish tissue using IMS (method C). Each method's performance was assessed according to the ISO standard requirements by testing shellfish homogenates seeded with different numbers of C. parvum oocysts. Two groups of parameters were compared, encompassing precision (coefficient of variation (CV)) and accuracy of measurements. These were described by linear regression models allowing calculation of the methods' limits of detection (LOD) and quantification (LOQ). In addition, oocyst recovery efficiencies from shellfish were calculated for each method. All three compared methods allowed for at least 66% recovery of Cryptosporidium oocysts from the tested samples. The best recovery (83.3–100%) in the whole range of tested suspensions was obtained for method C. The accuracy of method B was better (linearity of r2=0.9996 in the full measurement range) than that of method A (r2=0.968). Method C showed the best accuracy (r2=1) and precision (CV 0.2–14.1). Compared to other methods it was also characterised by the best LOD and LOQ, attaining ≅4 and ≅12 oocysts per 3 g of tested shellfish sample respectively. Despite a lack of the ability of method A to give the proportional results in oocysts recovery (non-linearity of the method) compared to the reference values, it achieved the highest LOD and LOQ values among the tested methods. As demonstrated here, the most efficient method for extraction of Cryptosporidium oocysts from shellfish tissues was method C employing sample homogenisation and separation of oocysts from tissue debris using IMS. Used alone this method does not in fact allow for identification of Cryptosporidium species but delivers quantitative results concerning the level of food contamination by parasites.