Pea cultivars are nearly homozygous and thus homogeneous when they are released. The traditional method of selfing is slow and inefficient, taking up to ten generations of inbreeding following a cross to achieve a high level of homozygosity. Current single-seed-descent (SSD) methodologies enable a maximum of three generations per year to be developed in pea. Doubled haploidy and an in vitro based modified SSD technology have been utilised in many important crops for the rapid achievement of homozygosity, and thus acceleration of the breeding process. In pea, due to the lack of robust protocols, none of these technologies is routinely used in a breeding program. The aim of this study was to accelerate the breeding process in pea by developing in vitro techniques to more rapidly achieve a high level of homozygosity and to gain a better understanding of the fundamental mechanisms involved in these processes. These techniques include: 1) haploidisation from cultured anthers in selected genotypes of varying backgrounds; and 2) in vitro flowering and seed-set for use in SSD breeding strategies. The development of robust genotype-independent in vitro protocols will be of great value to accelerate the breeding process in pea. In this research a number of key factors in the development of a robust pea anther culture protocol were identified and optimised. The combined application of multiple stress treatments, including the novel stress agent sonication, and the optimisation of key culture factors led to the development of an efficient protocol for the routine induction of androgenesis and embryo production from extracted anthers of various pea genotypes. A flow cytometry study was undertaken to further understand the effect of individual and combined stress treatments on androgenesis elicitation. Analysis of the flow cytometry results revealed clear differences in the relative nuclear DNA content of microspores within anthers after stress treatments iv and enabled prediction of whether a combination of stresses were elicitors or enhancers of androgenesis. Flow cytometry is thus proposed as a method for the quick assessment of the effect of individual and combined stress treatments, based on the relative nuclear DNA content. An optimised in vitro based SSD system was developed which enabled rapid in vitro flowering and seed-set across a range of pea genotypes including, for the first time, mid to late flowering types. In this protocol, the antigiberellin Flurprimidol was used to control in vitro plant size, and plants with the meristem removed and excised shoot tip explants were cultured into glass tubes under white fluorescent light. The involvement of antigibberellin and light quality on plant growth response is discussed. Using this strategy more than five generations per year can be obtained with mid to late flowering genotypes and over six generations per year for early to mid flowering genotypes. The results presented in this research form a solid basis for further efforts designed to enhance androgenic response in pea and extend double haploid technology to other legumes. However, further research is required before this technology will be routinely available within a pea breeding program. In the absence of a robust DH protocol, the in vitro based SSD system reported herein will offer a valuable alternative method for the rapid achievement of homozygosity by shortening each generation cycle.