Principal　Components　Analysis　(PCA),　Independent　Component　Analysis　(ICA)　and　Linear　Discriminant　Analysis　(LDA),　have　been　widely　used　for　2D　face　recognition.　Local　Binary　Pattern　(LBP),　however,　provides　a　simpler　and　more　effective　way　to　represent　faces.　With　LBP,　face　image　is　divided　into　small　regions　from　which　LBP　histograms　are　extracted　and　concatenated　into　a　single　and　global　feature　histogram　representing　the　face　image.　The　recognition　is　performed　using　Chi　square　and　other　commonly　used　dissimilarity　measures.　In　this　paper,　we　construct　LBP　codes　together　with　three　dissimilarity　measures　on　hexagonal　structure.　We　show　that　LBPs　defined　on　hexagonal　structure　will　lead　to　a　faster　and　more　accurate　scheme　for　face　recognition.