Most of the existing fingerprint indexing algorithms are based on either macro-level (level-1) details such as singularities or level-2 details such as minutiae in the fingerprint image. Level-3 features such as pores have not been explored much for fingerprint indexing as it requires fingerprint scanners with resolution greater than 1000 dpi. However, pores in fingerprint images are known to contain a significant amount of discriminatory information. Therefore, there is a need to investigate the effectiveness of pore features for fingerprint indexing. This paper presents a fingerprint indexing algorithm based on pore features which are extracted by applying Delaunay triangulation on the detected pores. Experiments are performed on the publicly available High Resolution Fingerprint (HRF) database (DBII). Performance measures from our experiments show the effectiveness of the proposed indexing algorithm and indicate that pore features alone are viable for fingerprint indexing.