The inheritance of yellow-podded versus green-podded peas was studied by Mendel more than 150 years ago, but, until now, the allelic variation underlying this difference remained unknown. A report of a novel, spontaneous gp mutant (Pellew and Sverdrup, 1923) suggested that at least two independent gp alleles existed. We undertook genetic complementation crosses with 19 yellow-podded lines present in the John Innes Pisum germplasm collection, one of which is described as “Pellew’s gp”, and we found that all were allelic to gp carried by the type-line, JI0128. This showed that Mendel’s yellow-podded character is conditioned by variation at a single locus. We used genetic mapping of Axiom markers in a large F2 population to define the position of gp within a 4.4 cM interval, encompassing approximately 8 Mb. We used exome capture to compare genome structures within this interval, in 227 Gp accessions and the 19 gp accessions. We found that all 19 gp lines carry the same large 100 kb deletion, while all 227 Gp accessions do not, indicating that only one gp haplotype exists in the germplasm we studied, and that gp must have been crossed into different backgrounds by breeders and researchers in the past. Three candidate genes associated with the deletion were assessed. After crossing a null TiLLING mutant for one of these candidates with a gp line, yellow-podded F1 progeny were obtained. This genetic complementation test showed that we have identified a new allele of the gp gene. We used backcross lines to gain a better understanding of the mutant phenotype. It was already known that thylakoid membranes are underdeveloped in gp lines (Price et al., 1988). Additional effects will be described.