The original vectors of the bacterial two-hybrid technique developed by Karimova et al. in 1998 did not enable detection of the recombinant proteins. Here, we propose two methods resolving this problem, either using new plasmids containing the Flag epitope, or using a trick to detect the T18 domain of adenylate cyclase. Furthermore, we describe a set of vectors for TAP, CBP or 6-histidine tagging that possess the same cloning site as our two-hybrid vectors.