AbstractSeveral multifactor crosses have been performed in which all segregating genotypes were identified. The genetic map of T4 and mapping techniques are re‐evaluated in light of the finding that double‐mutant recombinants occur less frequently than those identified as wild‐type. This difference is attributable to heterozygotes which exaggerate the latter class. Analysis of the crosses suggests that the mathematical mapping functions so far devised overcorrect for clustering of multiple crossovers, and in so doing predict more crossovers than actually occur. An 8‐factorrIIcross in which crossover estimates were based exclusively on double‐mutant recombinants shows that the high negative interference (HNI) observed in mass lysates is not an artifact of the method used to enumerate recombinants. Experiments to investigate the cause of HNI include parallel 8‐factorrIIcrosses between alternating point mutations and deletions. One cross was performed in the presence of inhibitory concentrations of fluorouracil deoxyriboside, the other in its absence. Based on double‐mutant recombination frequencies, neither case showed a difference in the coefficient of coincidence for insertion of a point mutant between two deletions and for insertion of a deletion between two point mutants. Either insertion heteroduplexes are not involved in formation of double crossovers, or deletion heteroduplexes (not observed among progeny phages) are efficiently repaired. Analysis of single‐burst progenies of cells infected with a partial phage genome and a helper phage shows a 6.8‐fold increase in recombination frequency near the ends of the partial phage genome. It is suggested that the HNI observed in mass lysates is a population genetics artifact accounted for by the increased frequency of recombination near ends