One problem associated with the development of subunit vaccines is their limited immunogenicity, due to their physico-chemical structure, their inability to encounter the correct MHC restriction element, or the need for strong adjuvants to be delivered along with them. These problems are usually solved by conjugating target epitopes (peptides or oligosaccharides) with carrier proteins which provide a source of T-cell epitopes recognised by a large proportion of the vaccinated individuals. We have shown that mycobacterial hsp65 and hsp70 exert a strong helper effect in vivo when conjugated to synthetic peptides or oligosaccharides. Interestingly, this helper effect did not require the need for any adjuvant, either in mice or in monkeys. The helper effect mediated by the hsp65 required that animals were previously primed with either live BCG or the hsp65 alone; on the other hand, such a priming was not required when the hsp70 was used in the conjugates. Similar results were obtained with HSP molecules from Escherichia coli. This may suggest that the adjuvant-free helper effect observed applies not only to mycobacterial HSP, but also to HSP from other prokaryotes. These findings suggest that microbial hsp70 could be considered for the design of conjugated vaccine constructs for eventual human use.