Phenotypic characterization and molecular evaluation were compared in their differentiating influence and quantity of information to assess the genetic diversity in guava. Forty nine genotypes were analyzed phenotypicaly using eleven quantitative traits; out of them ten were selected for the comparison. Analysis of variance showed highly significant differences amongst guava genotypes in all traits except total sugar content. On the other hand, the sequence related amplified polymorphism (SRAP) was used for the molecular analysis. A total of 88 SRAP amplicons were generated by five primer combinations, of which 58 bands (65.9%) were polymorphic. Both cluster analysis of unweighted pair-grouping method with arithmetic averages (UPGMA) and principal coordinate (PCO) analysis based on phenotypic data and SRAP clearly separated the genotypes according to their relationships. However, the clustering arrangement was different depending on the data used for the analysis. In addition, the percentage of polymorphism amongst the tested genotypes varied between the two marker systems. SRAP marker was able to generate some unique specific bands for certain genotypes which could be helpful for further use in guava genetic enhancement.