Abstract: A study of the combining ability of diallel crossing of Chinese fir was made in 1974 and 1975. The material for the tests came from a breeding population of plus tree clones, the ortets of which were selected in the Cunninghamia lanceolata plantations in Fujian. Mating plans for the factorial (1974, 1975), half diallel crossing (1975) and diallel crossing (1975) are shown in Figures 1-4. The main results of this study were obtained as follows:1. General combining ability (g. c. a. ), specific combining ability (s. c. a.) and reciprocal effects were highly significant as shown by the T-tests for height growth.2. Both the factorial and the half diallel crossing tests showed that height growth was under some g. c. a. control in C. lanceolata (h2Nl = 0.1 - 0.64, h2NF = 0.77 -- 0.86) The estimates for the g. c. a. of growth traits showed that there are dif: ferences as derived from various parents.3. Specific combining ability (s. c. a.) was also significant in both tests. Specific combining ability values also vary with different combinations of parents.4. As for the 4-year height growth, the tests showed obvious variations in trees of different parental origins, narrow-sense heritabilities of family (0-86) and individual trees (0-35) being very high. Reciprocal effects were significant at this stage in the diallel and half diallel tests, which indicated the presence of the dominant effect.5. The whole series of height measurements (from seedling to 4 yr.) in the diallel, half diallel and factorial tests indicated a progressive change in the formation of genetic control. Figures 8-12 show that s. c. a. effects count for most of the total genetic variance in the seedling stage (34-83%), but they steadily decline with years. On the other hand, the additive effects of g. c. a. steadily increase with years.6- Selfing effects were significant in the dialiel tests. The estimates of the effects of selfing on the height growth indicated that there were differences derived from parents.