Frequent itemset mining is one of the fundamental but time-demanding tasks in data mining. It is used to find frequent patterns and generate association rules for these patterns. With the availability of inexpensive storage and progress in data capture technology, the availability of data has reached exa-scale already. But improvements in processor and network technology open up opportunity for parallel and distributed computing to be applied in frequent itemset mining to improve its performance in the light of the challenge of “big data”. Thus, there are challenges in frequency itemset mining to fully harness the parallel processing capability of the computer hardware technologies. This paper reviews the development of current serial and parallel approaches to frequent itemset mining and discusses future research directions in this field.