SUMMARYIn order to determine how important gene conversion is as a force in populations, extensive surveys have been made to determine the limits, distributions and typical values of evolution-related conversion parameters from various fungi, Drosophila and maize. The conversion frequency, c, had an enormous range for different mutations within a species, with loci often having different means and limits for c. b, the frequency of a particular allelle in the products of meiotic tetrads or octads with aberrant segregation ratios, often showed values near the theoretical extremes; d, disparity in direction of conversion, was frequent and often extreme. b and d could only be studied in fungi, and their less extreme values in Saccharomyces than in Sordaria orAscobolus are clearly related to the lack of frame-shift mutations in the yeast data. For different mutations at a locus, neither c nor b showed normal distributions; both gave dispersed distributions, sometimes multi-modal for b; c and b were not usually correlated. y, the force of meiotic gene conversion on allele frequencies, had a large range of positive and negative values, with different loci often having quite different limits and mean absolute values. For different mutations at a locus, y showed dispersed distributions, with little tendency to cluster around zero. The high proportion of non-zero y values in all fungi where this could be studied, with large departures from zero being most extreme in Ascobolus and Sordaria brevicollis, shows that gene conversion could often be an important force in changing allele frequencies in favour of mutant or wild-type in these organisms. Even in organisms with much lower c or absolute d values, conversion could still be important, depending on dominance, selection coefficients and mutation rates. Presumed frame-shifts tended to have more extreme disparity and hence higher absolute y values than did presumed base-substitutions, but with no Consistent C differences. Induced mutations had similar or slightly higher absolute y values compared with spontaneous mutations.