Further studies on the use in peptide synthesis of N(π)-phenacyl protection for histidine side chains have shown that whilst this prevents the side chain-induced racemisation which can occur if there is a lone pair of electrons available at the π-nitrogen, there are concomitant drawbacks. As an alternative approach to the racemisation problem, the effect of halogenation of the heterocyclic ring carbons (to diminish the availability of the π-nitrogen lone pair) has been investigated. This gives derivatives which are convenient in both classical and solid-phase applications, the halogen modifying groups being removed at the last stage by catalytic hydrogenolysis over a rhodium catalyst. Racemisation is suppressed as expected, but it is not eliminated completely: direct blockade of the π-nitrogen appears to be indispensable for its complete prohibition. Protection of the π-nitrogen with a benzyloxymethyl group has now been found to be much more satisfactory than the use of the phenacyl group for this purpose. A π-benzyloxymethyl substituent not only prohibits side chain-induced racemisation but also gives derivatives with convenient physical properties which can be incorporated into well established classical and solid-phase strategies without the need for any novel or additional operations or changes in protocol. The protecting group is stable to basic conditions, to trifluoroacetic acid, and to aqueous solutions of carboxylic acids, but is cleaved cleanly and rapidly by hydrogen bromide in trifluoroacetic acid or by catalytic hydrogenolysis. N(α)-t-Butoxycarbonyl-N(π)-benzyloxymethyl-L-histidine has been prepared in good yield by a simple procedure from an easily accessible intermediate and isolated as a crystalline solid; its use has been demonstrated by a number of exercises including a solid-phase synthesis of 5-isoleucine-angiotensin II and a classical synthesis of trihistidine.