diff --git a/src/tokenize.py b/src/tokenize.py
index d4d60be..8286a0f 100644
--- a/src/tokenize.py
+++ b/src/tokenize.py
@@ -47,6 +47,7 @@
     SPACE = enum.auto()  # pragma: no mutate
     NEWLINE = enum.auto()  # pragma: no mutate
     BOOL = enum.auto()  # pragma: no mutate
+    KEYWORD = enum.auto()  # pragma: no mutate
 
 
 # Represents a tokenizer token
@@ -98,6 +99,24 @@
     return tokens
 
 
+# Keywords recognized by the language
+keywords = [
+    "NewLang",
+    "Done",
+    "Set",
+    "To",
+    "EndSet",
+    "If",
+    "Then",
+    "Else",
+    "EndIf",
+    "StartNote",
+    "EndNote",
+    "StartText",
+    "EndText",
+]
+
+
 # Classifies tokens in to types
 def classify_tokens(tokens):
     new_tokens = []
@@ -108,6 +127,8 @@
             type = TokenType.SPACE
         elif t.value in ["True", "False"]:
             type = TokenType.BOOL
+        elif t.value in keywords:
+            type = TokenType.KEYWORD
         else:
             type = TokenType.UNKNOWN
         new = Token(t.value, t.location, type)
diff --git a/tests/test_tokenize.py b/tests/test_tokenize.py
index 70c02ef..c7976d1 100644
--- a/tests/test_tokenize.py
+++ b/tests/test_tokenize.py
@@ -17,6 +17,24 @@
 from src import tokenize
 
 
+# Keywords recognized by the language
+keywords = [
+    "NewLang",
+    "Done",
+    "Set",
+    "To",
+    "EndSet",
+    "If",
+    "Then",
+    "Else",
+    "EndIf",
+    "StartNote",
+    "EndNote",
+    "StartText",
+    "EndText",
+]
+
+
 # Draws a random token location
 @composite
 def draw_token_location(draw):
@@ -49,6 +67,7 @@
     chars = characters(blacklist_characters=reserved)
     value = draw(text(alphabet=chars, min_size=1))
     assume(value not in ["True", "False"])
+    assume(value not in keywords)
     type = tokenize.TokenType.UNKNOWN
     return tokenize.Token(value, token.location, type)
 
@@ -84,6 +103,15 @@
     return tokenize.Token(value, token.location, type)
 
 
+# Draws a keyword token
+@composite
+def draw_token_keyword(draw):
+    token = draw(draw_token_random())
+    value = draw(sampled_from(keywords))
+    type = tokenize.TokenType.KEYWORD
+    return tokenize.Token(value, token.location, type)
+
+
 # Draws a classified token
 @composite
 def draw_token_classified(draw):
@@ -92,6 +120,7 @@
         draw_token_space(),
         draw_token_newline(),
         draw_token_bool(),
+        draw_token_keyword(),
     ]
     token = draw(one_of(strategies))
     return token