Thesis (MScAgric (Horticulture))--Stellenbosch University, 2008. / Chemical thinning of fruit trees has become a central management practice for ensuring high fruit quality at harvest and return bloom the following season. Three trials were conducted in the 2004/5, 2006/7 and 2007/8 seasons to investigate the efficacy and mode of action of chemical thinning agents on European pear cultivars (Pyrus communis L) in the Western Cape, South Africa. The first trial was conducted in the 2004/5 and 2006/7 seasons to evaluate the efficacy of 50, 100 and 150 mg.l-1 6-benzyladenine (BA), and 30 and 40 mg.l-1 naphthylacetamide (NAD) on ‘Early Bon Chrétien’ pear. BA was more effective than NAD in reducing crop load and improving fruit size. Crop load decreased and fruit size increased with increasing rate of BA. BA significantly improved, whilst NAD failed to improve return bloom. In the second trial, three experiments were conducted in the 2006/7 and 2007/8 seasons to evaluate the efficacy of 100 to 200 mg.l-1 BA on ‘Forelle’ pear. The first experiment was conducted in the 2006/7 season where BA rates of 100, 125 and 150 mg.l-1 generally failed to reduce crop load or to improve fruit size and fruit size distribution and return bloom. The second experiment was conducted in the 2007/8 season where two BA rates, 150 and 200 mg.l-1 and a split-application of 3 x 50 mg.l-1 improved fruit size. The 200 mg.l-1 rate was the most effective treatment. BA did not improve fruit size distribution and return bloom. The third experiment was conducted in the 2007/8 season where the effect of rate and timing of BA applications was evaluated. Two rates, 150 and 200 mg.l-1 were applied 8, 11 and 17 days after full bloom (d.a.f.b.). There was no significant interaction between BA rate and application time. The 200 mg.l-1 rate and the 11 d.a.f.b. (i.e. 8 to 10 mm average fruit size) applications were more effective in reducing crop load, and improving fruit size. BA at 150 and 200 mg.l-1 and at all application times significantly improved return bloom relative to the control. From these trials we concluded that BA is a reliable thinner for ‘Early Bon Chrétien’ at rates of 100 or 150 mg.l-1. On ‘Forelle’, BA is not a reliable thinner and we recommended further trials with BA in combination with other thinning agents. In the third trial, three experiments were conducted in the 2007/8 season to investigate the mode of action and effect of BA application time on European pear cultivars. The effect of site of application, bourse shoot growth and fruit size at time of application on the efficacy of BA was evaluated. Results from the experiments on the effect of site of application and bourse shoot growth were inconclusive. In terms of fruit abscission, there was a significant interaction between BA application time and fruitlet size. Early BA applications (8 d.a.f.b.) were significantly more effective in promoting fruit abscission, than later (11 and 17 d.a.f.b.) applications. Smaller fruit (6 to 8 mm) were found to be more susceptible to BA-induced fruit abscission than bigger fruit (8 to 12 mm).