Fifteen immunoassay kits for serum hCG(-β) measurements have been compared for their analytical characteristics and clinical usefulness in tumour monitoring. Three out of 13 radioimmunoassays and two sandwich hCG immunoassays represent the ‘one-component’ techniques for intact hCG detection since they are based on antiserum raised against intact hCG. Within this group, method-comparison slopes ranged between 0·149 and 0·374. The remaining 10 radioimmunoassays all apply antiserum against hCG-β and are designated ‘two (or more)–component’ assays since they detect both intact hCG and the free β-subunit (‘total hCG’). For this reason these assays are of potential value for tumour monitoring. Additional criteria for this are: (1) the expression of assay results in terms of 1st IRP-hCG instead of the unstructured use of the 2nd IS or the 1st IRP-hCG-β; (2) documentation of cross-reactivities especially for the free β-subunits of hCG and hLH which are almost completely lacking; (3) establishment of the minimum detectable dose in the presence of normo-to-hypergonadotrophic hLH levels in serum (the ‘clinical sensitivity’) to allow the follow-up of tumour regression especially in the low-dose region. Method-comparison analysis for these assays revealed regression slopes between 0·001 and 0·873.