A citrus spacing experiment begun in 1980 included 4 spacings, 2 varieties, and 2 rootstocks. Planting densities ranged from 150 to 360 trees per acre with spacings from 8 x 15 ft to 15 x 20 ft. 'Hamlin' and 'Valencia' Sweet orange (Citrus sinensis (L.) Osb.) trees were planted on Milam lemon ( C. jambhiri Lush) and Rusk citrange ( Poncirus trifoliata (l.) Raf. x C. sinensis) rootstocks. Trees at the 8 ft in row spacing grew together to form a continuous hedgerow in 5 yr and competition between trees reduced trunk growth. Yield increased with increasing tree density during the fourth and fifth seasons. 'Hamlin' trees produced more boxes per acre than 'Valencia' trees during the fourth and fifth seasons. Trees on Rusk citrange rootstock yielded substantially better during these seasons than trees on Milam. The planting was subiected to freezes each year since planting. Trees at closer spacings received less freeze damage during the last 2 winters. A limited supply of acreage suitable for citrus, high land values, increasing property taxes, and high production costs, coupled with periodic depressed fruit prices make a high per acre production an economic necessity. Higher density plantings have offered early production in spacing experiments outside Florida (1, 2, 4, 8, 10) and in a Florida experiment with 'Pineapple' orange (5, 6). The concept of higher density plantings has been reviewed (11), and trends toward closer tree spacings (9) have continued to the present time. However, optimal management practices, spacing, scion and stock combinations, and their effects on tree size and productivity have not been established. The purpose of this report is to describe an experiment begun in 1980 and to report results for the first 5 seasons after planting. The objectives of the experiment were: (a) to determine whether higher density plantings can be managed to realize early production and returns without creating future problems requiring excessive pruning and/or tree removal; (b) to develop production, harvesting, and fruit handling practices which can be utilized in the efficient management of such plantings; (c) to evaluate water and nutrient requirements relative to spacing; and (d) to determine optimal tree spacing and height to maximize profitability for the scion-stock combinations used in this study. The information presented here includes effects of scion, rootstock, and spacing on tree growth, mineral nutrition, fruit yield and quality, and freeze damage to trees from 1980 to 1986. Florida Agricultural Experiment Station Journal Series No. 7783. The contributions and cooperation of the Coca-Cola Company Foods Division in making this experiment possible are gratefully acknowledged. Proc. Fla. State Hort. Soc. 99: 1986. Materials and Methods Factors included in this experiment are listed in Table 1. A multiple split-plot experimental design was used with 4 replications. Scion variety was the main plot followed by smaller subplots of tree height, between-row spacing, rootstock, and in-row spa_..cing in that order. The in-row and between-row spacings listed in Table 1 are presented in the text, tables, and figures rounded to the nearest foot; i.e., 8 x 15 ft instead of 8.2 x 14.8 ft spacing. Plot size was 4 rows x 6 trees with the center 10 trees (2 rows x 5 trees) used for collecting data. The appropriate analysis of variance model was used for statistical analysis of all data, and any treatment differences presented were significant at the 1 or 5% level. The experiment is located on a 25-acre site in Polk County, FL, between the towns of Babson Park and Frostproof. A permanent overhead sprinkler irrigation system to provide uniform application was installed prior to planting the trees. 'Hamlin' and 'Valencia' were selected as scion varieties to represent early and late varieties and because they were not included in the previous major Florida spacing experiment (5, 6). Milam and Rusk citrange were selected as a vigorous and moderately vigorous rootstock, respectively. Including several scion/stock combinations permits evaluation of the horticultural adaptability of various combinations to the constraints of higher density plantings. Tree height treatments have not been started as trees have not yet reached the required heights. Trees for this experiment were grown in a commercial nursery. Trees were headed out at a 24 inch height which resulted in a tree slightly larger than usual at the time of planting. A heading height higher than normal was used to reduce the number of low branches to facilitate future mechanical harvesting or fruit handling studies. Trees were planted at the experimental site in February, 1980 and a regular commercial young tree care program was followed for the first 3 yr, followed by commercial grove cultural practices. Trunk circumference measurements were made annually 8 inches above the ground. Beginning in the fall of 1984, leaf samples were collected annually for mineral Table 1. Experimental factors and tree density.