In clonus and progenies linear phenotypic and genotypic correlations and interactions between genotypes, years and locations as well as methods for the estimation of the heritability are investigated.The significant correlation coefficients found should facilitate selection for more than one character at one time. With one exception high genotypic correlations correspond to significant phenotypic ones (Table 1). Usually genotype/year and genotype/year/location-interactions were not significant, but genotype/location-interactions were regularly significant in trials carried over more than one season. There were also clear differences in the ability of genotypes to yield consistently in different locations over several seasons ("ecovalence").Estimates of heritabilities both in the narrow sense and in the broad sense were relatively high, especially those of flowering date, plant height and leaf number. Both types of heritabilities showed the same ranking (Table 3). Half-sib families from polycrosses were not suitable for half-sib correlations.From the results obtained it has been concluded that one year trials with replicated clones in different locations yield useful estimates of heritability. Therefore, in breeding plants with long generation times, raising of progenies may be dispensed with in the interest of saving time.