Purification methods for Cryptosporidium oocysts are usually selected on the basis of recovery yield, but the amount of particulate debris in environmental matrices could limit efficiency of oocyst detection by microscopic examination or PCR detection. Previous studies have shown that the standard immunomagnetic separation (IMS) procedure would not be the most suitable method for oocyst purification from turbid matrices. We compared the capacity of Percoll-sucrose flotation and six other density-based purification methods to achieve selective separation of Cryptosporidium oocysts from particulate debris. Rate of oocyst recovery and particulate loading in the purified suspensions were chosen as comparison criteria for the different purification methods. In most earlier studies, the chemical treatments employed to obtain a purified oocyst suspension modify the surface properties of oocysts in spiked samples. Assuming this produces unrealistic conditions affecting the evaluation of purification methods, we performed the present study with native oocysts. Flotation and gradient procedures were tested with and without formaldehyde ethyl acetate (FEA) separation. FEA separation was found to be unsuitable. Filtration and Percoll gradient did not allow selective oocyst separation from debris. Among the purification methods suitable for routine microscopic examination, Percoll-sucrose flotation provided the best recovery rates. For automated enumeration systems or PCR detection, potassium bromide and especially Nycodenz gradients appeared to be the most suitable purification methods. Potassium bromide and Nycodenz gradients provided the best balance between oocyst recovery and particulate load.