Monthly observation of cocoa flushing, number of cherelle wilt (CW), number of small, medium and large pods of 6 clones was conducted for two years to study its dynamics for one year. A study was conducted in Kaliwining Experimental Station, 45 m asl. and D rainfall type (according to Schmidt & Ferguson), using ICS 13, ICS 60, TSH 858, Sulawesi 1, Sulawesi 2 and KW 165 clones of 8 years old. Each clone was planted intermittently in separate rows, replicated 6 rows. Correlation and regression analysis were done between variables and with rainfall data. The parallel research was conducted in the similar station to assess the accuracy of production estimation method by identify percentage of small pods (length 1—10 cm), medium (11—15 cm) and large pods (>15 cm) to grow until harvested. The study used 15th years old trees of Sulawesi 1, Sulawesi 2, KW 165, KKM 22, ICS 13 and DR 2 clones. Each clones was replicated 5 times. The result showed that intensive flushing (>50%) occured during January, March, September and November meanwhile no flushing during December and February. Correlation between rainfall and flushing was positive (r=0.27). Effect of clones on flushing frequency was similar but for flushing intensity was significant. KW 165 tended to be the lowest but TSH 858 tend to be the highest. CW occured for a year-round but the height level during May and June. Effect of clones was significant, KW 165 showed highest followed by Sulawesi 2. CW level showed positive correlation with number of medium (r=0.71) and big pods (r=0.55), except showed negative correlation with flushing intensity (r=-0.37) and rainfall (r=-0.51). High pod setting happened during May to November and low pod setting during December to March. In this aspect effect of clones were significant, the productive clones were Sulawesi 1, Sulawesi 2 and KW 165, but ICS 60 was the less. CW level during 1st semester was lower than at 2nd semester and clone effect was significant. The opportunity of small, medium and big pods to be harvested was similar among both semester. The opportunity of small, medium and big pods to be harvested were 8—56%; 57—83% and 77—96% respectively depend on the clones. In average, those opportunities were 27%, 72%, and 87% for small, medium and big pods respectively.Key words: flushing, cherelle wilt, pod setting, clones, yield prediction