Rab GTPases are post-translationally modified by addition of geranylgeranyl moieties to carboxyl-terminal cysteine residues. For Rab proteins ending with xxCC xCxC and CCxx motifs this modification is catalysed by geranylgeranyltransferase type II (GGTaseII), and is entirely dependent on the Rab substrate being bound to Rab escort protein (REP). Several Rab proteins contain carboxyl-terminal CaaL prenylation motifs typical of members of the Rho family, which are modified in a REP-independent manner by geranylgeranyltransferase type I (GGTaseI). The present studies show that one such Rab protein (Rab8), which ends with a CVLL motif, is uniquely able to serve as a substrate for either REP/GGTaseII or GGTaseI in cell-free assays. The modification of Rab8 by GGTaseI did not require REP, indicating that a REP-induced conformational change is not essential for exposure of the Rab carboxyl-terminal cysteine prenylation site. To determine whether one enzyme plays a predominant role in Rab8 prenylation in vivo, the incorporation of [3H]mevalonate into Rab8 was measured in human embryonal kidney 293 cells under conditions where the activity of GGTaseI, but not GGTaseII, was blocked by the peptidomimetic inhibitor GGTI-298. The GGTaseI inhibitor did not prevent prenylation of either overexpressed Myc-tagged Rab8 or endogenous Rab8, whereas prenylation of a known GGTaseI substrate with the same carboxyl-terminal motif, Cdc42Hs, was completely blocked. To rule out the possibility that the apparent prenylation of Rab8 by GGTaseII occurs only when GGTaseI activity is eliminated, metabolic labelling studies were carried out in the absence of the GGTaseI inhibitor, using a REP-binding-deficient Rab8 construct (Y78D) that cannot serve as a substrate for GGTaseII, but is indistinguishable from wild-type Rab8 as a substrate for GGTaseI. Prenylation of the Y78D mutant was reduced by 60-70% in intact cells, consistent with the conclusion that the majority of Rab8 is prenylated by the REP/GGTaseII system in vivo.