Whilst exogenous cytokinins have been shown to delay the rapid post-harvest senescence of broccoli florets (Brassica olerácea var. Itálica), there has been little research to date to examine, in depth, the rôle endogenous cytokinins may play throughout the degradative process. If attempts are made to delay senescence via the manipulation of cytokinins using genetic engineering techniques, then it is important to gain some understanding of the major cytokinins in the floret tissue. This includes determining how the cytokinin concentrations and metabolic capacity change over time, and identifying the significance of degradative enzymes that affect the active cytokinin pool. In order to study changes in the cytokinin profiles throughout senescence, tissue samples for cytokinin analyses were taken from freshly harvested florets (cv. “Green Belt”), as well as from florets which had been stored at either 5°C for up to 24 days, or 20°C for 6 days. Within each temperature treatment, some florets were stored under 18 hrs of light and others were kept in darkness, and all were provided with 95% relative humidity under conditions which prevented ethylene accumulation. Broccoli which senesced rapidly, such as that kept at 20°C, or in darkness, generally had lower concentrations of zeatin, zeatin riboside, dihydrozeatin and dihydrozeatin riboside, isopentenyl adenine and isopentenyl adenosine than florets subjected to light or 5°C storage conditions. Furthermore, the concentrations of these cytokinins were always greater in florets stored at 5°C compared with 20°C, regardless of light treatment. The concentrations of the cytokinin bases were usually greater than their riboside form. Comparisons of the same cytokinin types were made between two broccoli cultivars, which had previously been identified as having either good or poor post-harvest storage capabilities. All cytokinin types were greater in the cultivar with superior storage capabilities (“Green Belt”) compared with the inferior cultivar (“Green Beauty”) after the florets had been subjected to dark storage at 20°C. Again, cytokinin bases were present in greater concentrations than the ribosides. When stems of stored florets were placed in [14C]-zeatin or [14C]-benzyladenine for 6 hours, cytokinin was taken up and rapidly converted within that time to the riboside form, adenine, or a highly polar unknown compound. With advanced senescence (Le. florets stored for 6 days at 20°C or 24 days at 5°C), up to 100% of uptake was metabolised into other forms within 6 hours. The conversion of [14C]-zeatin into adenine by cytokinin oxidase decreased with advancing senescence in “Green Belt” florets stored at 5°C, but increased with senescence at 20°C. Cytokinin oxidase activity decreased in “Green Beauty” florets as they senesced, regardless of storage temperature. Regardless of storage temperature, the majority of zeatin taken up by the florets was converted to the riboside form with increasing senescence in both broccoli cultivars, and the total active cytokinin concentration (zeatin and zeatin riboside) was always greater in florets stored under light than in darkness. Florets of both cultivars stored under light at 5°C maintained greater [14C]- benzyladenine concentrations than florets kept in darkness, but the opposite occurred in florets stored at 20°C. Rapid metabolism occurred even with advanced senescence, with up to 70% of uptake being converted to an unknown polar compound during the 6 hours of the cytokinin feed. In vitro cytokinin oxidase studies using [14C]-zeatin as the enzyme substrate revealed that the optimum reaction temperature was 37°C, optimum pH of 7.0, and Vmax and Km values, respectively, of 37.04 pmol.mg protein-1.hr-1 and 1.64 μM. Reactions using tritiated compounds showed the enzyme had a greater affinity for zeatin, followed by zeatin riboside, isopentenyl adenosine and isopentenyl adenine. Enzyme purification (using a 40 - 70% ammonium sulphate precipitation cut instead of 0 - 80%) resulted in a smaller protein yield but also a 4-fold increase in the amount of zeatin conversion to adenine. The amount of zeatin converted to adenine (under optimal reaction conditions) was greater (per gram of original tissue) in florets subjected to light than darkness at 5°C, and greater in “Green Beauty” florets compared with “Green Belt” within each light treatment. A plant regeneration protocol using tissue culture techniques was established specifically for the “Green Belt” broccoli cultivar preparatory to genetic transformation using a particle inflow gun. Attempts were made to introduce into the broccoli genome a binary vector containing a senescence-retarding gene (the ipt gene from Agrobacterium tumefaciens) attached to a senescence-associated gene promoter. Suitable expiants for this work included peduncles and hypocotyls, but not petioles. Optimum shooting protocols have also been described. At this stage, however, no stable transformants have been recovered, although transient GUS expression was observed.