Concerns have been raised about the reliability of dot-probe tasks. The cued Visual Probe Task (cVPT) uses cues predicting locations of emotional stimuli, which appears to improve reliability. However, cVPT reliability could be affected by individual differences involving cue features. Here, we assessed specifically anticipatory reliability. Further, trial-to-trial carryover effects, previously found for stimulus-evoked biases, were tested. 82 participants were analysed, who performed an online procedure including a reversal of the cue mapping. Predicted stimulus categories were neutral and angry faces. Cue-Stimulus Intervals of 400 and 1000 ms were used. An overall anticipatory attentional bias, in terms of RT difference scores, towards threat was found. Reliability was around .4, similar to previous results despite the mapping reversal procedure. Carryover effects were found with a similar pattern as for non-cued threat-evoked bias. The results confirm a reasonably reliable outcome-focused bias towards threat, showing similar carryover effects as found for stimulus-evoked bias.