The effects of ethephon on arabica coffee (Coffea arabica L.) were tested in experiments conducted over 2 years in North Queensland. Rates tested were 0, 125, 250, 500, 750, 1000, 1500, and 2000 mgkg, applied when 15-20% (early treatment) or 45-50% (late treatment) of fruit were coloured. Measurements were made on the degree of leaf and fruit abscission, berry ripening, and berry weight. Ethephon did not cause significant fruit abscission, but rates >500 mg/kg caused unacceptable leaf abscission. Rates of 250 mg/kg were marginally acceptable in terms of leaf abscission. Excessive defoliation led to dieback, which resulted in a reduced crop the following season. Ethephon concentrated berry ripening, significantly increased the number of red (mature) berries, and decreased the number of green (immature) berries. Rates of >500 mg/kg were effective in accelerating berry ripening, while 125 mg/kg had a limited effect. Maturity stage of the bean at time of application helped to determine the ethephon response; early application of ethephon had a more pronounced effect on berry ripening than late application. Ethephon applied early at rates of 1500 and 2000 mg/kg decreased fresh berry weight compared with all other treatments but parchment weight was unaffected. Selection of a suitable rate of ethephon from these data is difficult because berry ripening is offset by leaf drop. Variability of response was also observed.