A number of different approaches have been used for genotyping hepatitis C virus (HCV). Two simplified methods were evaluated, both of which used polymerase chain reaction (PCR) to amplify products from the 5Π non-coding region of HCV: non-isotopic restriction fragment length polymorphism (RFLP) analysis and type-specific PCR. Sixty-four viraemic patients suffering from chronic HCV infection were studied using these two techniques; 25/64 samples were further tested with a commercial serotyping ELISA based on synthetic NS4 antigen (Murex, U.K.). The results of the three typing methods were generally in agreement with each other. When only the predominant genotype identified by each method was analysed, the 3 methods had 100% agreement. RFLP did not detect any mixed infections and it was unsuccessful in 16/64 (25%) samples. Both type-specific PCR and serotyping ELISA detected mixed infections. However, serotyping ELISA did not give typeable results in 7/25 (28%) samples, whereas type-specific PCR gave typeable results in all 64 samples. Typespecific PCR detected more mixed infections than serotyping ELISA. Direct sequencing of four PCR products with indeterminate RFLP confirmed changes in restriction enzyme recognition sites. The sequences also confirmed the validity of the predominant genotype in cases of apparent mixed infections. It is possible that some of these cases were artefacts as a result of quasispecies.