It is questionable as to whether a low serum concentration of one of the IgG subclasses identi es a disease state. A low IgG 1 concentration is found in primary or secondary immunode ciency states but does not occur in isolation. Low IgG 2 concentration is associated with an increased risk of bacterial infections but only in some individuals and not in others. Isolated IgG 3 and IgG 4 de ciency have not been convincingly demonstrated. Therefore, the isolated nding of low concentrations of one or more IgG subclass does not identify individuals at risk. In contrast, the nding of low serum concentrations of antibodies to speci c bacterial antigens (Haemophilus in uenzae type B, pneumococcus, tetanus and diphtheria) does identify individuals at risk and these measurements should be used in preference to IgG subclass measurement.