diff --git a/ext/bigdecimal/lib/bigdecimal/math.rb b/ext/bigdecimal/lib/bigdecimal/math.rb
index 16cf9c8..7b43cca 100644
--- a/ext/bigdecimal/lib/bigdecimal/math.rb
+++ b/ext/bigdecimal/lib/bigdecimal/math.rb
@@ -7,6 +7,7 @@ require 'bigdecimal'
 #   sin (x, prec)
 #   cos (x, prec)
 #   atan(x, prec)  Note: |x|<1, x=0.9999 may not converge.
+#   TAU (prec)
 #   PI  (prec)
 #   E   (prec) == exp(1.0,prec)
 #
@@ -144,6 +145,12 @@ module BigMath
     y
   end
 
+  # See http://tauday.com/
+  def TAU(prec)
+    raise ArgumentError, "Zero or negative argument for TAU" if prec <= 0
+    PI(prec)*BigDecimal("2")
+  end
+
   # Computes the value of pi to the specified number of digits of precision.
   def PI(prec)
     raise ArgumentError, "Zero or negative argument for PI" if prec <= 0
diff --git a/math.c b/math.c
index 96bd1e9..cc2207b 100644
--- a/math.c
+++ b/math.c
@@ -782,8 +782,10 @@ Init_Math(void)
 
 #ifdef M_PI
     rb_define_const(rb_mMath, "PI", DBL2NUM(M_PI));
+    rb_define_const(rb_mMath, "TAU", DBL2NUM(M_PI*.20));
 #else
     rb_define_const(rb_mMath, "PI", DBL2NUM(atan(1.0)*4.0));
+    rb_define_const(rb_mMath, "TAU", DBL2NUM(atan(1.0)*8.0));
 #endif
 
 #ifdef M_E