The philosophical basis and utility of DNA barcoding have been a subject of numerous debates. While most literature embraces it, some studies continue to question its use in dipterans, butterflies and marine gastropods. Here, we explore the utility of DNA barcoding in identifying spider species that vary in taxonomic affiliation, morphological diagnosibility and geographic distribution. Our first test searched for a ‘barcoding gap’ by comparing intra- and interspecific means, medians and overlap in more than 75 000 computed Kimura 2-parameter (K2P) genetic distances in three families. Our second test compared K2P distances of congeneric species with high vs. low morphological distinctness in 20 genera of 11 families. Our third test explored the effect of enlarging geographical sampling area at a continental scale on genetic variability in DNA barcodes within 20 species of nine families. Our results generally point towards a high utility of DNA barcodes in identifying spider species. However, the size of the barcoding gap strongly depends on taxonomic groups and practices. It is becoming critical to define the barcoding gap statistically more consistently and to document its variation over taxonomic scales. Our results support models of independent patterns of morphological and molecular evolution by showing that DNA barcodes are effective in species identification regardless of their morphological diagnosibility. We also show that DNA barcodes represent an effective tool for identifying spider species over geographic scales, yet their variation contains useful biogeographic information.