diff --git a/c64/basic/basic_main_loop.s b/c64/basic/basic_main_loop.s
index 03b673af..719ed601 100644
--- a/c64/basic/basic_main_loop.s
+++ b/c64/basic/basic_main_loop.s
@@ -36,13 +36,58 @@ got_line_of_input:
 	ldx #$00
 	jsr pack_word
 
-	;; Print the packed word
-	lda #<$0100
+	;; Search for word in keyword list
+	;; The packed form is canonical, so we can search for the keyword while
+	;; it remains compressed. This also saves a little time.
+	;; Algorithm is simply to find matching first bytes, and then look for
+	;; the rest being a match.
+
+token_search:
+	;; tokenise_work1 = offset in line of input
+	;; tokenise_work2 = length of token we are matching
+	;; tokenise_work3 = offset in token list
+	ldy #$00
+	sta tokenise_work1
+	sta tokenise_work3
+next_kw_offset:	
+	;; Advance offset in compressed keyword list, to see if a match
+	;; BEGINS here
+
+	;;  Get count of bytes to compare
+	lda tokenise_work2
+	sta tokenise_work4
+	sta $0427
+	
+	;; Load current position
+	ldx tokenise_work1
+	ldy tokenise_work3
+	
+	;; Advance offset in compressed token list
+	inc tokenise_work3
+	lda tokenise_work3
+	cmp #$ff
+	bne next_in_match
+	jmp done_searching_for_token
+next_in_match:	
+	lda $0100,x
+	cmp packed_keywords,y
+	bne next_kw_offset
+	;; Advance to next chars in source and target
+	inx
+	iny
+	dec tokenise_work4 	; Have we compared all bytes yet?
+	bne next_in_match
+	;; Keyword matches!
+	inc $d020
+	tya
+	clc
+	adc #<packed_keywords
 	sta temp_string_ptr+0
-	lda #>$0100
+	lda #>packed_keywords
 	sta temp_string_ptr+1
 	jsr print_packed_string
 	
+done_searching_for_token:	
 	jsr ready_message
 
 	jmp basic_main_loop