Combined action observation and motor imagery (AO + MI) can improve movement execution (ME) in healthy adults and certain patient populations. However, it is unclear how the specificity of the observation component during AO + MI influences ME. As generalised observation could result in more flexible AO + MI rehabilitation programmes, this study investigated whether observing typing of target words (specific condition) or non-matching words (general condition) during AO + MI would have different effects on keyboard typing in healthy young adults. In Experiment 1, 51 students imagined typing a target word while watching typing videos that were either specific to the target word or general. There were no differences in typing execution between AO + MI conditions, though participants typed more slowly after both AO + MI conditions compared with no observation or imagery. Experiment 2 repeated Experiment 1 in 20 students, but with a faster stimulus speed in the AO + MI conditions and increased cognitive difficulty in the control condition. The results showed that the slowed typing after AO + MI was likely due to a strong influence of task-switching between imagery and execution, as well as an automatic imitation effect. Both experiments demonstrate that general and specific AO + MI comparably affect ME. In addition, slower ME following both AO + MI and a challenging cognitive task provides support for the motor-cognitive model of MI.