The pyrimidine-3 mutants of Neurospora crassa have received considerable attention in the past. Mitchell and Mitchell' described experiments, using pyr-3a (37301), 3b (37815; temperature sensitive), and 3d (45502), which indicated that the pyr-3a and 3b mutations represent a genetic unit distinct from 3d. The evidence for this view was that pyrimidine-independent heterocaryons could be formed by the combination of pyr-3a or 3b with 3d; that 3d, unlike 3a or 3b, was unaffected by a suppressor, s; and that crosses between 3a or 3b and 3d gave rise to a few (less than 0.01 per cent) wild type progeny. The segregation of a closely linked colonial marker (co), however, suggested that the wild type progeny arose through a mechanism different from simple reciprocal chromosome exchange during meiosis.Suyama et al.2 have confirmed and extended these data with a large series of independently isolated pyr-3 mutants. Their work indicates that the pyr-3 locus consists, on the basis of heterocaryon complementation, of two functional units, and that a large portion of the wild type progeny arising from intercrosses of mutants can be ascribed to a mechanism ("gene conversion") which does not involve the recombination of closely linked genetic markers. All of the pyr-3 mutants, including both functional units revealed by heterocaryon complementation, appear to be arranged in a continuous, linear order, and reasonable additivity of distances prevails in the genetic map. The maximum percentage of prototrophs from crosses of mutants was approximately 0.02 per cent, which indicated a very short length of the pyr-3 complex. (The pyr-3d mutant is reported to be identical with a mutant, KS36, in one of the functional units described by Suyama et al.;2 pyr-3a and 3b presumably represent the other unit.)Because the patterns of heterocaryon complementation or suppressor relationship observed among pyr-3 mutants has been observed in other mutant series concerned with one36 or with more than one enzyme, it cannot be decided with the evidence available whether or not all pyr-3 mutations affect a single metabolic function.This investigation was designed to relate enzymatic features of the pyr-3 mutants to the implied physiological differences between them and to the genetic structure of the pyr-3 locus. Preliminary data have indicated that pyr-3a, 3b, and 3d, like the wild type strain, all contain the series of enzymes required for the transformation of ureido-succinic acid (US; = carbamyl aspartic acid) to uridylic acid via the orotic acid pathway, and that deficiencies in the activity of certain of these enzymes are responsible for the pyrimidine requirement of other pyr mutants.7 US does not satisfy the nutritional requirement of any of the pyr-3 mutants. The preliminary enzymatic studies, however, were considered compelling evidence that the metabolic block in pyr-3 mutants concerned the synthesis of US or earlier steps of pyrimidine metabolism.8 Attention was therefore directed to the synthesis of 677