In order to improve the activity of hammerhead ribozymes in vivo, we have analyzed the effect of several prototypical RNA binding proteins on the ribozyme cleavage reaction: bacteriophage T4 gene 32 protein (gp32), hnRNP Al (Al) and the nucleocapsid protein of HIV-1 (NCp7). We show that, while gp32 has no effect on the cleavage reaction, Al and NCp7 affect different steps of the reaction. Moreover, some of these effects depend upon the ribozyme - substrate hybrid length. A1 and NCp7 inhibit the reaction of the least stable ribozyme-substrate complexes, which have 12 bp of duplex. NCp7, but not Al, inhibits the cleavage of substrates that have long ribozyme-substrate duplexes (17 or 20 bp), while cleavage of complexes having shorter duplexes (13 or 14 bp) is not affected. NCp7 and Al enhance the turnover of ribozymes by increasing the rate of product dissociation, but only when both cleavage products are bound with less than or equal to 7 bp. Al and NCp7 enhance ribozyme binding to long substrates, such as mRNAs, the structure of which otherwise limits ribozyme binding. Therefore, the effects of Al or NCp7 on the different steps of the cleavage reaction define a length of the ribozyme-substrate duplex which allows enhancement of the rate of binding and product release without inhibiting the cleavage step. Interestingly, this duplex length (14 bases, or 7 on each side of the cleavage site) is identical for Al and NCp7. Since Al is thought to interact with most, if not all mRNAs in vivo, it may enhance the intracellular activity of ribozymes targeted against any mRNA. On the other hand, since retroviral nucleocapsid (NC) proteins interact only with the viral genomic RNA in vivo, these results raise the exciting possibility that NCp7 and other NC proteins may specifically enhance the activity of ribozymes targeted against their cognate retroviruses. Overexpression of Al or NCp7, and targeting the ribozyme to a cellular compartment rich in Al or NCp7 may also enhance ribozyme activity in vivo.