Through his network of correspondents, Crépin gradually built up a huge herbarium of roses, which he would later donate to the State Botanic Garden in Brussels.
Toward the end of his career, his herbarium of roses included more than 40,000 collections, representing the whole natural range of the genus Rosa. Additionally, he received thousands of sheets on loan from botanists and botanical institutions all over the world, which he had carefully studied.
In all, he estimated that he had received some 100,000 collections for revision and comment, which were afterwards returned to their collectors or institutions, who valued the authoritative comments he added on the labels. He was also asked to verify and comment on exsiccatae prepared for distribution among interested researchers and institutions. These collections included Herbarium Rosarum (published by the Association rhodologique française and edited by the French botanists Simon Pons and Hippolyte Coste) and collections from the Botanical Exchange Club of the British Isles. Subscribers also received booklets with additional critical notes on each collection in the exsiccatae. Cuttings from these booklets were often added to the sheets with the dried specimen.
Strongly averse to speculation, Crépin preferred observation to hypothesis. As an evolutionist, he knew that understanding the deep geological past was important for defining the relationship between species and varieties existing today and building a classification reflecting natural affinities. Crépin distinguished between book species created and published by phytographers (those concerned with the detailed description of plants) and natural species. Whereas book species were contentious creations of the human mind, Crépin was convinced that natural species really existed ‘out there’ and were not merely human creations - he was a ‘species realist’.
Crépin maintained that species were essentially different from genera or varieties, which are creations of the human mind. Here, however, lay the key question he struggled with throughout his career: how to define and recognise a natural species? And how to avoid confusing a natural species with a genus that contains multiple species with a mere variety lacking sufficient distinguishing features to make it a ‘good’ separate species?