This thesis reports the establishment of protoplast isolation, culture and regeneration protocols, as well as cell selection systems and the optimisation of electroporation parameters for direct gene transfer to asparagus (Asparagus officinalis L.). Shoots of in vitro cultures were chosen as the source material to minimise the genetic instability associated with the use of callus in previous studies. A number of elite genotypes were used in this study, most of which are under evaluation as the parents of new hybrid cultivars. A protocol to maximise the release, isolation and purification of viable protoplasts from shoot cultures of asparagus was established. Protoplasts were purified by suspending digested etiolated shoot tissue in 0.6 M sucrose, overlaying with KMG medium (KM medium with 0.6 M glucose) and centrifuging at 650 x g. Important factors for high yield of viable protoplasts included: the use of in vitro etiolated shoots as source material; 0.6 M glucose as an osmoticum in modified KM medium; a combination of pectinase, cellulase and hemicellulase, each at 1 % (w/v) for enzymatic digestion of cell walls; as well as physical factors such as the volume of enzyme solution and speed of gyratory shaking. The asparagus genotype had a marked influence on protoplast yield, with some genotypes yielding up to 18.4 x 10??? protoplasts/g fresh etiolated shoot tissue with 90% viability. An optimized protocol was developed for the culture and regeneration of asparagus protoplasts. In the presence of actively growing asparagus feeder cells, asparagus protoplasts from etiolated shoots initiated cell divisions within 2 days of embedding in agarose beads to give final plating efficiencies of up to 20%. About 80% of the protoplasts initiating cell division developed into cell colonies, of which 25 % initiated somatic embryos. Complete plants have been regenerated from the embryos at a frequency of over 20% and have maintained phenotypically normal growth for over two years under greenhouse conditions. Chromosome counts confirmed the diploid status (2n = 2x = 20) of plants regenerated via somatic embryogenesis. Overall this protocol resulted in regeneration of complete plants from about 0.8% of the initially isolated protoplasts. Based on the established protoplast culture system, a range of antibiotics and herbicides were surveyed for their toxicity to protoplasts in order to develop a cell selection system for asparagus transformation. Asparagus protoplasts responded to the various selective agents in different manners. The low toxicity of some herbicides was attributed to the exogenous supply of amino acids as an essential component of the culture medium. A few chemicals were found to have poor repeatability in their toxicity response, while others consistently inhibited the growth of asparagus protoplasts at defined concentrations. Geneticin (G418) was identified as the preferred selective agent due to high effectiveness at low dose rates, a narrow concentration range from no effect to total inhibition and high repeatability of toxic response between experiments. A reconstruction experiment designed to evaluate the effectiveness of this selection system established a cross-feeding effect between non- transgenic and transgenic asparagus cells expressing a NPTII gene, resulting in a significant proportion of "escapes" at geneticin concentrations normally expected to be inhibitory to their growth. An increase in the concentration of the selective agent could eliminate this cross-feeding effect. Since initial attempts at transient ??-glucuronidase expression in electroporated asparagus protoplasts were unsuccessful, Nicotiana plumbaginijolia Viv. protoplasts as a positive control and biolistic transformation of callus of tobacco and asparagus and small onion bulbs were used to isolate the possible causes of the negative results. These experiments confirmed the effectiveness of DNA, the electroporation unit, and the competence of asparagus for direct gene transfer. Therefore, further experiments were conducted with asparagus protoplasts using an exponential electroporation machine to investigate the electrical parameters important in electroporation. These experiments demonstrated the key effect of energy dissipation in electroporation. There exists a good linear relationship between energy dissipation and both electroporation efficiency (EE) and survival rate. Time constant, which reflects the waveform of energy dissipation, has a minor influence on EE. At the same energy level, longer time constants were more effective at increasing EE. Results demonstrated that energy dissipation per unit volume of protoplast suspension should also be considered in electroporation experiments. A theoretical basis for these new concepts in electroporation was derived using energy as the key parameter. This principle should be generally applicable to electroporation of other species, and is not machine dependent. The application of this new principle was successful in achieving transient expression of a ??-glucuronidase gene following electroporation of asparagus protoplasts. In conclusion, the results and findings from this thesis demonstrated that direct gene transfer to asparagus based on the protoplast system is feasible, and protocols established in this thesis provide an opportunity for the genetic manipulation of asparagus using protoplasts.