Guava is cultivated in Bangladesh, India, Thailand, Brazil, Florida, West Indies, California and many other countries of the world. Worldwide Pakistan is the 2 nd largest producer after India. In Pakistan it ranks fifth. It is cultivated under the area of 72 thousand ha and production is 546 thousand tons. It is an allogamous crop (25-40% cross pollination) which causes the genetic divergence, on one hand it causes clonal degradation and on other hand it add new cultivars. There are different types of breeding methods to improve the crop but our focus on selection. We used 132 promising guava (Psidium guajava L.) genotypes collected from 12 areas in Pakistan. A total of 33 traits (18 qualitative and 15 quantitative) were applied to assess the genetic variability and structure of these guava germplasm. Characters including fruit acidity, fruit diameter, seed weight, non-reducing sugar, thickness of outer flesh, number of seed, fruit sweetness, longitudinal grooves, leaf twisting, fruit color of skin, fruit shape at stalk, longitudinal ridges and color of flesh were highly variable. Many of these traits in this study are of high economic importance and can be used as breeding targets on fruit yield and quality. Strong positive correlations were detected among 15 quantitative traits related to fruit yield and quality including fruit length and fruit diameter, fruit weight and fruit diameter, length of leaf blade and width of leaf blade, number of seeds and seed weight, fruit weight and diameter of cavity on fruit, seed weight and fruit weight. But there were more negative correlations among 18 qualitative traits studied. These 132 accessions involving 33 traits were categorized by principle components analysis (PCA). Results on correlation relationship among these traits were also projected in 2D plot. The morphological dendrogram generated from agglomeration hierarchical clustering (AHC) can distinguish all the 132 accessions into 3 major clusters. Images of genetic diversity and structure on these guava accessions were also presented in 2D plot. In this paper we report, for the first time, genetic variability and structure of Pakistan guava germplasm and its implications on conservation and breeding. Our results showed that morphological traits (using XLSTAT software) were very appropriate for studying genetic relationships among guava accessions and can be used for other open pollinated fruit plant germplasm management and breeding strategies as well.Retrotransposon-based DNA marker system (iPBS) and micro-satellite (SSR) markers were used to assess molecular variation and genetic structure in 51 promising Pakistani guava genotypes Psidium guajava L. together with 19 international accessions. The main objective of this research was to compare results derived from data sets for varietal identification and diversity estimation in Guava germplasm by iPBS and SSR marker systems. PCR from 5 iPBS primers (dominant markers) produced a total of 90 bands (100% polymorphic ) ranging from 100 bp to 2600 bp, and the mean PIC value for each primer ranging from 0.2339 to 0.3698. The 18 pairs of SSR primers (co-dominant markers) produced a total of 71 bands (100% polymorphic) ranging from 150 bp to 900 bp, and the mean PIC value for each primer ranging from 0.0848 to 0.9891. Molecular information generated from both iPBS and SSR methods were separately scored in a matrix for phylogenetic dendrogram construction as well as principle components analysis (PCA). Both iPBS and SSR markers can discriminate all the accessions evaluated into diversity groups. The phylogenetic dendrogram based on iPBS markers reflected not only the botanical morphologic classifications of the plants studied, but also the geographic distribution of some major international guava cultivars (Pakistan, Mexican, Hawaiian and Indian). In SplitTree results from iPBS method, 51 Pakistan accessions of Psidium guajava from 6 districts were distinguished into 6 groups according to their original locality. Results from the molecular analysis correspond strongly to that of PCA. Comparing to the SSR marker system, the iPBS PCR-based genome fingerprinting technology in this study is low-cost and provides another effective alternative in differentiation accessions in guava (Psidium guajava Linn.) and related species or genera. These results showed that iPBS marker system was very appropriate for study of genetic relationships among guava accessions and can be used for study of the genetic structure of Myrtaceae germplasm as well as other plants.