We carried out a field experiment in order to study effects of fertilization in juvenile plants of three coffee (Coffea arabica) cultivars in Yunnan, SW China. Fertilization treatments included a control without fertilizer (CK), combinations of three NPK fertilization rates [high fertilization (F H ), medium fertilization (F M ), and low fertilization (F L ) with 135, 90, and 45 g per plant per year, respectively], and at two N:P 2 O 5 :K 2 O ratios (R 1 , 1:0.5:0.8; R 2 , 1:0.8:0.5). The growth in juvenile plants was not altered by fertilization, with two clear growth peaks being observed in both the height and stem growth rates (RGRs) throughout a year. Both F M and F H resulted in significantly higher RGRs in both height and stem diameter compared to F L and CK in all three cultivars. At the same fertilization rate, the leaf area, branch number, longest branch length, internode number, and biomass of R 2 were higher than those of R 1 , and P significantly affected the root biomass and root to shoot ratio. Compared to the F L treatment, both F M and F H treatments resulted in higher net photosynthetic rates and stomatal conductance across seasons, and in higher intrinsic water-use efficiency during the dry season and at the middle of the wet season. Photosynthetic nitrogen use efficiency at R 2 was higher than at R 1 , but no significant differences were observed among the different fertilization rates. Among the three coffee cultivars, Caturra exhibited the highest height, stem diameter, longest branch length, and internode number. Our results indicated that the optimal N:P 2 O 5 :K 2 O ratio is 1:0.8:0.5 for the juvenile growth of coffee plants. Both F M and F H could help optimize the growth and photosynthetic rate of coffee plants, but F M is suitable for the ecological friendly agriculture and economic sustainability in coffee plantations.