Previous studies demonstrated that redundant target stimuli can enhance performance due to multisensory interplay and interactively facilitate performance enhancements due to temporal expectations (TE; faster and accurate reactions to temporally expected targets). Here we tested whether other types of multisensory interactions – i.e. interactions evoked by temporally flanking irrelevant stimuli – can result in similar performance patterns and boost not only unisensory target perception (multi– vs. unisensory sequences) but also unisensory temporal expectations (expected vs. unexpected). To test our hypothesis, we presented sequences of 12 stimuli (10 Hz) which either consisted of auditory (A), visual (V) or alternating auditory–visual stimuli (e.g. A–V–A–V–etc) with either auditory (AV(A)) or visual (AV(V)) targets. Participants had to discriminate target frequency which was unpredictable by temporal regularities (expected vs. unexpected target positions) and by stimulation sequence (A, V, AV(A), AV(V)). Moreover, we ran two experiments in which we presented redundant multisensory targets and manipulated the speed of the stimulation sequence (10 vs. 15 Hz stimulus trains) to control whether the results of Experiment 1 depended on sequence speed. Performance for unisensory targets was affected by temporally flanking distractors, with multisensory interactions selectively improving unisensory visual target perception. Yet, only redundant multisensory targets reliably affected TEs. Together, these results indicate that cross–modal facilitation of unisensory target perception in fast stimulus streams is uni–directional, but also differs from multisensory interactions evoked by redundant targets; more specifically, it appears to be context–dependent (task, design etc.) whether unisensory stimulation (unlike redundant target stimulation) allows for the generation of temporal expectations.