Empirical observations carried out by the first author during his many visits to cocoa producing countries suggested that a relationship might exist between cocoa pulp flavour attributes and sensory traits of cocoa liquor and chocolates. A first attempt to demonstrate such a relationship was made by evaluating the following flavour attributes: volatile aroma upon opening the pods, pulp sweetness, pulp acidity, pulp astringency, pulp flavour intensity and type, overall preference for pulp taste, and degree of bitterness of the cotyledon. Blind pulp tasting was carried out by a six-person panel of ripe fruits of nine cocoa clonal varieties, representing different genetic origins, several of which are well known for their contrasting sensory traits of their chocolates. Overall pulp preference was significantly related to the known fine flavour potential of the cocoa varieties tested, whereas specific pulp flavour attributes appeared related to specific types of fine flavour of the tested cocoa genotypes. One striking example was the EET 62 clone (Nacional x Trinitario), which displayed a "jasmine" flower odour upon opening the ripe fruit, sweet pulp with an intensive pleasant flavour that was identified as floral and fruity. These flavour attributes are apparently related to the known typical "Arriba" flavour of the EET62 clone. The typical fresh-fruit Trinitario cocoa flavour of the ICS 1 clone appeared to be related to a balanced combination of relatively high acidity and sweetness and a moderate citrus-like flavour of its pulp. On the other side, the pulp of the potentially low-flavour CCN51 clone was rated as very astringent and acid, with low flavour intensity and quality. The pulp of the Pandora 1 clone from Colombia was characterised by a sour sop (Guanabana) flavour and may therefore represent a new interesting type of fine-flavour cocoa. These results suggest that a significant part of fine-flavour attributes of cocoa products (excluding basic cocoa flavour) c