Phenol is an important intermediate for the production of petrochemicals, agrochemicals and plastics. Most phenol worldwide is produced via the three-step cumene process from benzene. The disadvantages of the cumene process have stimulated the development of alternative routes to decrease energy consumption, increase the yields and avoid the explosive intermediate (cumene hydroperoxide) and the byproduct acetone. The one-step direct hydroxylation of benzene toward phenol has been extensively investigated using various oxidants such as molecular oxygen, nitrous oxide and hydrogen peroxide. But the selective production of phenol with high yields remains a challenge because phenol is more reactive toward oxidation than benzene. It is difficult to achieve high selectivity using molecular oxygen because of the high reaction temperatures and overoxidation which arise from the more active catalysts used for activating oxygen. The application of nitrous oxide is limited by the source, although the phenol selectivity and conversion can be high. Amongst the oxidants, hydrogen peroxide has obvious advantages as water is the only byproduct. In the past several decades, a variety of catalysts have been developed for the direct hydroxylation of benzene with hydrogen peroxide as the oxidant. Catalysts based on molecular sieves show outstanding catalytic performances for the reaction. This perspective article discusses the direct hydroxylation of benzene to phenol using hydrogen peroxide with catalysts based on molecular sieves, including pristine molecular sieve catalysts, transition metals incorporated molecular sieves, transition metal oxides supported on molecular sieves, complexes grafted on molecular sieves, and heteropoly acids supported on molecular sieves.