As summarized in this minireview, two different signal recognition events, one involving SRP and the other involving proteoliposomes containing the Sec61p complex, have been identified. In cotranslational protein transport, it seems that both recognition events are required for efficient translocation of the protein into the lumen of the ER. The requirement for SRP can, under certain experimental conditions, be circumvented by depletion of NAC, a heterodimeric complex that can block the tight association of nascent chain-ribosome complexes to the Sec61p complex in the ER membrane. In posttranslational protein transport, the Sec61p complex contains additional protein subunits that are required for function. It should be noted that, in all the experiments performed in which the role of SRP in cotranslational protein translocation is circumvented (Jungnickel and Rapoport, 1995; Lauring et al., 1995a, 1995b), stable translocation intermediates are allowed many minutes to establish productive interactions with the membrane. In contrast, during conditions in which the nascent chain can elongate (e.g., in vivo), the nascent chain-ribosome complex only has a brief time window during which it can initiate translocation (reviewed by Walter and Johnson, 1994). It is possible that, under these conditions, productive translocation even in the absence of NAC would require SRP. The isolation of NAC-deficient extracts that support protein synthesis will allow a test of this possibility. Finally, the role that lipids themselves may play in protein transport should not be ignored. Gierasch and coworkers (Hoyt and Gierasch, 1991, and references therein) have shown that bacterial signal peptides have an intrinsic ability to interact with lipid and that the relative ability of a mutant signal sequence to interact with lipid correlates with its function as a signal sequence in vivo. Thus, the signal sequence-discriminatory role defined by Jungnickel and Rapoport (1995) may in fact be played by lipid, with the Sec61p complex playing a necessary but nondiscriminatory role in the process. In this light, it is interesting that Martoglio et al. (1995) recently demonstrated that the signal sequence of preprolactin could be cross-linked to phospholipid. Analysis of the cross-linking efficiency of the signal sequence to phospholipid at different nascent chain lengths and with mutant signal sequences will help define the role that phospholipid plays in the process.