The relationship between complete deficiency for either isotype of the fourth component of complement, C4A or C4B, and glomerulonephritis was initially examined in white patients from Kentucky with either IgA nephropathy or Schönlein-Henoch purpura. Subsequently, C4B deficiency was found to be associated with IgA nephropathy for pediatric patients followed in Cincinnati, Ohio. We later reported that at least 60% of the original patients from Kentucky were related to at least one other patient with the disease. This finding raised the possibility that the C4 phenotype frequencies for these patients may have been biased by the fact that they were based on a sample of related patients. In our study, C4 phenotyping was performed for 52 related and 63 unrelated patients from Kentucky, 81 unrelated patients from the Mid-South region of the United States, and 39 unrelated patients from the Puglia region of southeastern Italy. In addition, data from patients with IgA nephropathy from Spain were available for comparative studies. Neither C4A deficiency nor C4B deficiency was significantly increased for groups of unrelated patients from the Mid-South, Italy, Spain, or Kentucky in comparison with regional control subjects. In fact, C4A and C4B deficiencies did not occur in any of the Italian patienis. With the exception of C4A*6, frequencies for the most common C4A and C4B alleles did not differ among the unrelated patient and control groups from Kentucky and the Mid-South. In addition, no significant differences in C4A and C4B allelic frequencies were observed in comparisons of pediatric patients (diagnostic biopsy before age 18 years) and adult patients with IgA nephropathy in either U.S. population. Italian control subjects had a significantly lower frequency for C4A null alleles in comparison with control subjects from both Kentucky and the Mid-South; a significantly higher frequency of C4B null alleles was found among Kentucky control subjects than in Mid-South, Italian, and Spanish control samples. The importance of recognizing the ethnic background of study subjects and of eliciting a good family history to minimize unsuspected sampling of related patients should be considered in future disease association studies. © 1990 Mosby-Year Book, Inc.