The scarcity of feed resources often imposes a major challenge to the development of animal production in the tropics and subtropics. However, by-products have long been recognised in providing cheaper alternative feed ingredients relative to conventional feed ingredients, thus alleviating the challenge. The aim of the present study was to investigate the potential of three waste products from the subtropical fruit processing industry in animal feeding. The products include avocado meal (AM), macadamia oil cake (MOC) and macadamia chips (MCH). The samples were obtained from the processing plants in Nelspruit of Mpumalanga Province, Makhado (Louis Trichardt) and Tzaneen of Limpompo Province. Eight samples of each waste product were collected, prepared and their chemical composition, in vitro dry matter digestibility and, in situ ruminal dry matter and crude protein (CP) degradability were determined. A broiler growth trial was conducted as well to determine the effect of replacing maize with AM on performance of broilers under commercial production thus establishing the replacement value of avocado meal in broiler feeds. As the products under study were from the oil extraction process of the two fruits, there were high remnants of oil in all three waste products. As a result the ether extract (EE) content of the three waste products was higher compared to any of the commonly used oilseed meals. The MCH had the highest EE concentration, followed by the MOC and the AM had the lowest. However, after defatting the AM had the highest (P<0.05) EE concentration compared to either the MOC or MCH. The CP concentration of the MOC was significantly higher than that of the AM and MCH. There was no significant difference observed between the AM and MCH in CP concentration. A better amino acid profile was observed with the MOC compared to the AM and MCH. None of the products can be regarded as a protein source. The fibre fractions of the MCH were significantly higher than that of AM and MOC, with the exception of the acid detergent lignin (ADL) concentration. The ADL concentration of the AM and MCH did not differ significantly although the MCH concentration was higher. The MOC had the lowest (P<0.05) ADL concentration compared among these waste products. Generally, the fibre concentration in the MOC was lower compared to other waste products but higher compared to the SBM and POCM. The acid detergent insoluble nitrogen (ADIN) concentration of the AM was significantly higher than that of MOC and MCH. There was no significant difference observed between MOC and MCH. The condensed tannin (CT) concentration of the waste products differed significantly with the AM being the highest and the MOC the lowest. The ADIN concentration of the MOC could be compared to that of the SBM and lower than that of the POCM. There were some significant differences observed in the mineral composition between the waste products under study. The mineral concentrations were below the maximum tolerable levels of animals except for iron (Fe) in AM which can be toxic to sheep as it was above the maximum tolerable level of 500mg/kg. The in vitro dry matter digestibility of the three waste products differed significantly, with the MOC being the highest and the MCH the lowest. Huge variation within waste products was observed and it could be due to the contamination levels of the products with the indigestible portions of the parent fruits. The in situ degradability of dry matter and CP of the AM and MOC differed significantly, with the MOC surpassing the AM in most of the degradability characteristics. The potential degradable fraction (“b”) and the degradation rate of the b fraction (“c”) fractions of the AM and MOC did not differ significantly. The MCH had more indigestible particles of the kernel and as a result it could not be analysed statistically and was therefore omitted. The effect of replacing maize with AM at different inclusion rates led to decreased feed intake and the final mass of the broilers during the trial period. The feed intake of the broilers on commercial diet was significantly higher than that of the broilers on commercial diets with avocado meal, except for the inclusion rate of 10% AM. The final mass, the ADG and the FCE of the broilers on commercial diet were significantly higher compared to the broilers on diet with AM irrespective of the inclusion rate. No mortalities observed during the experimental period. Copyright / Dissertation (MSc(Agric))--University of Pretoria, 2012. / Animal and Wildlife Sciences / unrestricted