The objetive of this experiment was to study the adaptation of both tree species to agro-ecological conditions and to determine the fruit yield reducing factors. Soil type was a stony andisol. The sour guava experimental plot, consisted of ungrafted trees (15 years old average). Phenological studies have beenconducted and production indexes determined during the last three years. Pest incidence on buds, flowers and fruits was also observed. The heaviest blooming was in March and July, with little variation from year to year. Fruit production was concentrated in March, November and December. All trees reduced fruit production during the dry season. Fruit set varied from 0 to 12% depending on the season and individual tree. The average fruit production was 30 kg/tree, but in better quality soils around the same area yields reached 40 kg/tree. A sample of 200 fruits was evaluated and classified as follows: first class 14%, rejects 26% (because of overripening, fungal infections, trips, crakings and presence of fruit flies' larvae. The spanish plum, "tronador" type, plot was three and one half years old. They were propagated by cuttings, and planted 4x4 meters, and pruned twice a year. Their branches were ringed. The trees bloomed at the end of March, after a defoliation period whieh lasted until the end of May. Prunings at the begining of Marchdelayed blooming and fruit set. Flowers were produced in axilar panicles in variable numbers. Each panicle set one to three fruits. First harvest was at their third year and fruit numbers fluctuated from 20 to 140 fruits per tree. They were pruned during the last-quarter moon. The first two years' shape prunnings were done as tip cuttings. The spanish plum produced more secondary branches when tips were cut oof or branches ringed, and the resulting small branches were the productive ones. The foliage was usually pest free, except for so me aphids and worms. Fruits showed a heavy attack of unknown trip and mite species.