Recent studies reported that training of working memory may improve performance in the trained function and beyond. Other executive functions, however, have been rarely or not yet systematically examined. The aim of this study was to test the effectiveness of inhibitory control (IC) training to produce true training-related function improvements in a sample of 122 healthy adults using a randomized, double-blind pretest/posttest/follow-up design. Two groups performed either adaptive (training group) or nonadaptive (active control) versions of go/no-go and stop-signal tasks for 3 weeks. Training gains as well as near-transfer to an untrained Stroop task and far-transfer to psychometric fluid intelligence were explored. Although the adaptive group could substantially improve overall IC task performance after training, no differences to the active control group occurred, neither at posttest nor at follow-up testing. A large decrease in response latency from pre- to posttest (and from pretest to 4 months' follow-up testing) was found when the training group was compared to the passive control group, which, however, does not sufficiently control for possible confounds. Thus, no conclusive evidence was found that this performance increase mirrors a true increase in IC function. The fact that training improvement was mainly related to response latency may indicate that individuals were more focused on performance gains in the prepotent go trials but less on the stop trials to meet the requirements of the tasks as well as possible. The challenges for response inhibition training studies are extensively discussed.