Described is a procedure for serum C-reactive protein (CRP) determination, consisting of a semiquantitative rapid CRP latex agglutination test, using dilutions of the serum, and the quantitating spot immunoprecipitate assay (SIA). These methods are performed with standard laboratory equipment using no more than 30 microliters of serum for both assays. With visual inspection, the SIA results are available one to two hours after blood sampling. CRP levels obtained by agglutination testing with five batches of latex reagents coated with rabbit anti-CRP agree well with the quantification, i.e. essentially 100% for the negative sera and about 80% for positive sera containing greater than 40 mg CRP/l. The remaining 20% of the samples are classified as low positive at 10 to 40 mg/l on agglutination. False positive or negative agglutination findings are below 2.3% with concordance at 88% between SIA and the CRP agglutination with rabbit antibodies. Two lots of CRP-latex reagents coated with sheep antibodies, however, gave 15.3% and 10.1% false positive findings and poor concordance with SIA ratings, particularly for low positive sera at only 20 and 29%. SIA is suggested for CRP quantification because it compares well with radial immunodiffusion in accuracy (less than 91%) and provides results in 2 h rather than 1-2 days. Rocket electroimmunoassay is less reliable with lower ratings than found in SIA, probably due to the electrophoretic heterogeneity of CRP. This is demonstrated for two of three purified CRP preparations, for which varying agglutination is seen. The combination of methods is especially recommended for diagnosis and monitoring of CRP in infectious processes in neonates and infants because of the required small sample volume--0.5 ml heel-prick blood--the rapidity of reliable (greater than 80%) reporting and the possibility of rating sera with moderate levels of CRP.