Abstract The mitochondrial intron rI1 is a self-splicing group-II intron of algal mitochondria that can be transferred into chloroplasts from the green alga Chlamydomonas reinhardtii for in vivo investigations (Herdenberger et al. 1994). Thus, rI1 is a suitable system to compare in vitro and in vivo RNA processing. Interestingly, rI1 shows correct RNA splicing, although typical cis-acting exon-sequences (IBS2, ) of group-II introns are lacking. In order to examine the effect of these exon-intron interactions on splicing, we introduced the endogenous mitochondrial IBS2 sequence in order to produce optimal IBS2-EBS2 base pairing. In addition, the first nucleotide of the 3exon () was substituted to create an optimal - interaction. Neither of the two mutations, nor a combination of both, had any effect on the precision of the splice-site selection. Unexpectedly, introduction of IBS2 led to a reduction in the efficiency of the second splicing step in vitro but not in vivo. These findings lead us to conclude that trans-acting factors are present in vivo to optimize splicing efficiency. The possibility is discussed that these factors may, for example, stabilize tertiary intron structures that are a prerequisite for correct RNA processing. Furthermore, our data indicate that similar trans-acting factors promote correct intron splicing in chloroplasts and mitochondria.