diff --git a/S32-num/modf.t b/S32-num/modf.t
new file mode 100644
index 0000000000..8fbe59ba80
--- /dev/null
+++ b/S32-num/modf.t
@@ -0,0 +1,305 @@
+use Test;
+# L<S32::Numeric/Numeric/"=item modf">
+
+#plan 244;
+#plan 4;
+#plan 20;
+
+my $debug = 0;
+
+# test values and data defined in the BEGIN block below
+my @n;
+my $i;
+
+sub test-routine($val, $exp-int, $exp-frac, , $places, :$test-sign!) {
+}
+sub test-method($val, $exp-int, $exp-frac, , $places, :$test-sign!) {
+}
+
+# test the internal subs with positive data
+$i = 0;
+for @n -> ($val is copy, $exp-int is copy, $exp-frac is copy, $places) {
+    ++$i;
+
+    # raw data are originally Cool strings with leading minus signs
+    my $test-sign = +1;
+    test-helpers($val, $exp-int, $exp-frac, , $places, :$test-sign!) {
+}
+
+# test the internal subs with negative data
+$i = 0;
+for @n -> ($val is copy, $exp-int is copy, $exp-frac is copy, $places) {
+    ++$i;
+    my $test-sign = -1;
+    test-helpers($val, $exp-int, $exp-frac, , $places, :$test-sign!) {
+}
+
+
+
+=begin comment
+
+next;
+
+    note "DEBUG: item $i val = '$val', int = '$Int'. frac = '$Frac', places = '$places'" if $debug;
+    my $s = _num2str $val;
+    cmp-ok $s, '~~', Str;
+    note "DEBUG: item $i val = '$val', _num2str = '$s', str2num '{$s.Num}'" if $debug;
+
+#    next;
+
+    my $I = $Int.Str;
+    my $F = $Frac.Str;
+    my $P = $places.Str;
+
+    cmp-ok $val, '~~', Str;
+    cmp-ok $I, '~~', Str;
+    cmp-ok $F, '~~', Str;
+    cmp-ok $P, '~~', Str;
+
+    $val = $val.Real;
+    if $val ~~ Rational {
+        $val = $val.base(10, *);
+    }
+    else {
+        $val = $val.base(10, $places);
+    }
+
+    my ($left, $right) = _modf-str $val;
+    note "DEBUG: val = '$val', base(10) = '$left'" if $debug;
+    is $left, $I;
+    is $right, $F;
+
+}
+=end comment
+
+done-testing;
+
+=begin comment
+# test the routine
+#for @n -> ($val is copy, $Int is copy, $Frac is copy, $places) {
+for %n.kv -> $val is copy, $arr {
+    my $int-part  = @($arr)[0];
+    my $frac-part = @($arr)[1];
+    my $places    = @($arr)[2];
+
+    # check the negative values
+    my ($int, $frac) = modf $val;
+    is $int, $int-part, "modf($val), int: expected $int-part, got $int";
+    is-approx $frac, $frac-part, "modf($val), approx frac: expected $frac-part, got $frac";
+
+    # using places for exact frac match
+    ($int, $frac) = modf $val, $places;
+    is $frac, $frac-part, "modf($val, $places), exact frac: expected $frac-part, got $frac";
+
+    # check the positive values
+    $val       .= abs;
+    $int-part  .= abs;
+    $frac-part .= abs;
+
+    ($int, $frac) = modf $val;
+    is $int, $int-part, "modf($val), int: expected $int-part, got $int";
+    is-approx $frac, $frac-part, "modf($val), approx frac: expected $frac-part, got $frac";
+
+    # using places for exact frac match
+    ($int, $frac) = modf $val, $places;
+    is $frac, $frac-part, "modf($val, $places), exact frac: expected $frac-part, got $frac";
+}
+
+# test the method
+#for @n -> ($val is copy, $Int is copy, $Frac is copy, $places) {
+for %n.kv -> $val is copy, $arr {
+    $val .= Real;
+
+    my $int-part  = @($arr)[0];
+    my $frac-part = @($arr)[1];
+    my $places    = @($arr)[2];
+
+    # check the negative values
+    my ($int, $frac) = $val.modf;
+    is $int, $int-part, "$val.modf(), int: expected $int-part, got $int";
+    is-approx $frac, $frac-part, "modf($val), approx frac: expected $frac-part, got $frac";
+
+    # using places for exact frac match
+    ($int, $frac) = $val.modf($places);
+    is $frac, $frac-part, "$val.modf($places), exact frac: expected $frac-part, got $frac";
+
+    # check the positive values
+    $val       .= abs;
+    $int-part  .= abs;
+    $frac-part .= abs;
+
+    ($int, $frac) = $val.modf;
+    is $int, $int-part, "$val.modf(), int: expected $int-part, got $int";
+    is-approx $frac, $frac-part, "modf($val), approx frac: expected $frac-part, got $frac";
+
+    # using places for exact frac match
+    ($int, $frac) = $val.modf($places);
+    is $frac, $frac-part, "$val.modf($places), exact frac: expected $frac-part, got $frac";
+}
+=end comment
+
+sub test-helpers($val, $exp-int, $exp-frac, , $places, :$test-sign!) {
+    
+}
+
+sub test-routine($val, $exp-int, $exp-frac, , $places, :$test-sign!) {
+}
+sub test-method($val, $exp-int, $exp-frac, , $places, :$test-sign!) {
+}
+
+sub _num2str(Cool $n is copy --> Str) {
+    #===========================================================
+    # THIS SUB IS DUPLICATED IN CORE AND ROAST AND MUST REMAIN
+    # IDENTICAL IN BOTH PLACES. THE TWO FILES CONTAINING IT ARE:
+    #   repo https://github.com/rakudo/rakudo/src/core.*/Num.pm6
+    #   repo https://github.com/Raku/roast/S32-Num/modf.t
+    #===========================================================
+
+    # the incoming value a string representing a number
+    my $s = $n.Str;
+    given $s {
+        when /^ <[+-]>? \d* '.' \d* $/ {
+            # base 10 rational as string
+            # use as is
+        }
+        when /^ :i <[+-]>? 0 <[box]> <[abcde\d]>* '.' <[abcde\d]>* $/ {
+            # base 2, 8, or 16 rational as string
+            # use base 10,*
+            $s = $s.Num.base(10, *)
+        }
+        when /:i e/ {
+            # 1.3e-1
+            # use base, but how many places?
+            $s = $s.Num.base(10, 15)
+        }
+        when /:i '/' / {
+            # 1/3
+            # rational as string
+            # use base 10,*
+            $s = $s.Num.base(10, *)
+        }
+        when $s.Numeric ~~ Int  {
+            $s = $s.Num.base(10)
+        }
+        default {
+            note "|--  DEBUG: incoming, unhandled value '$n' is type '{$n.^name}'";
+        }
+    }
+    $s
+}
+
+sub _modf-str(Str $s is copy, $places? --> List) {
+    #===========================================================
+    # THIS SUB IS DUPLICATED IN CORE AND ROAST AND MUST REMAIN
+    # IDENTICAL IN BOTH PLACES. THE TWO FILES CONTAINING IT ARE:
+    #   repo https://github.com/rakudo/rakudo/src/core.*/Num.pm6
+    #   repo https://github.com/Raku/roast/S32-Num/modf.t
+    #===========================================================
+
+    # the incoming string was a number converted to
+    # an integer OR a decimal number
+
+    # check for the sign, if any
+    my $sign = '';
+    if $s ~~ /^ (<[+-]>) (.*) $/ {
+        $sign = ~$0;
+        $s    = ~$1;
+    }
+
+    my ($left, $right);
+    my $idx = index $s, '.';
+    if $idx.defined {
+        $left  = $s.substr: 0, $idx;
+        $right = $s.substr: $idx+1; # do NOT keep the decimal point here
+
+        # get the length of the bare right side
+        my $nr-chars = $right.chars;
+        # now embellish it
+        $right = '0.' ~ $right;
+
+        if $sign {
+            $left  = $sign ~ $left;
+            $right = $sign ~ $right;
+        }
+        # pad right with zeroes if need be
+        # note we do NOT trim any chars if right is too long
+        if $places.defined and $places > $nr-chars {
+            my $n = $places - $nr-chars;
+            $right ~= ('0' x $n);
+        }
+        $right = _trim-zeroes $right;
+
+        $left  = '0' if $left.Num == 0;
+        $right = '0' if $right.Num == 0;
+    }
+    else {
+        $left = $s;
+        note "DEBUG dotless left: '$left'" if $debug;
+        $left  = $sign ?? ($sign ~ $left) !! $left;
+        $left  = '0' if $left.Num == 0;
+        $right = '0';
+    }
+    $left, $right
+} # sub _modf-str
+
+sub _trim-zeroes(Str $x is copy --> Str) {
+    #===========================================================
+    # THIS SUB IS DUPLICATED IN CORE AND ROAST AND MUST REMAIN
+    # IDENTICAL IN BOTH PLACES. THE TWO FILES CONTAINING IT ARE:
+    #   repo https://github.com/rakudo/rakudo/src/core.*/Num.pm6
+    #   repo https://github.com/Raku/roast/S32-Num/modf.t
+    #===========================================================
+     
+    # the incoming string may have a decimal point. if
+    # so, leave it
+    while $x ~~ /<[.1..9]>0$/ {
+        note "DEBUG before trim '$x'";
+        $x ~~ s/0$//;
+        note "DEBUG  after trim '$x'";
+    }
+    $x
+} # sub _trim-zeroes
+
+BEGIN {
+
+    @n =
+    # 27 elements
+
+    #=== two cases reported by reviewers:
+    # possible problem child:
+    ['-6.2'                 , -6,                    -0.2000000000_0000000000_0000000000_0000000000_0000000000_0000000000_0000000000_0000000000_0000000000_0000000000, 100],
+    # okay tests
+    ['-123456789123456789.6', -123456789123456789,   -0.6,              1],
+
+    #=== other possible problem children:
+    ['-1.5e-10'             ,  0,                    -0.000_000_000_15, 11],
+    ['-123457e-3'           , -123,                  -0.457,            3],
+
+    # okay tests
+    ['-0x10.1'              , -16,                   -0.0625,           4],
+    ['-0o10.1'              , -8,                    -0.125,            3],
+    ['-0xa.1'               , -10,                   -0.0625,           4],
+    ['-0b10.1'              , -2,                    -0.5,              1],
+    ['-0b10'                , -2,                     0,                0],
+    ['-10'                  , -10,                    0,                0],
+    ['-0x10'                , -16,                    0,                0],
+    ['-0o10'                , -8,                     0,                0],
+    ['-0xa'                 , -10,                    0,                0],
+    ['-0o127'               , -87,                    0,                0],
+    ['-100'                 , -100,                   0,                0],
+    ['-5.9'                 , -5,                    -0.9,              1],
+    ['-5.499'               , -5,                    -0.499,            3],
+    ['-2'                   , -2,                     0,                0],
+    ['-3/2'                 , -1,                    -0.5,              1],
+    ['-1.5e0'               , -1,                    -0.5,              1],
+    ['-1.4999'              , -1,                    -0.4999,           4],
+    ['-1.23456'             , -1,                    -0.23456,          5],
+    ['-1'                   , -1,                     0,                0],
+    ['-0.5'                 ,  0,                    -0.5,              1],
+    ['-0.499'               ,  0,                    -0.499,            3],
+    ['-0.1'                 ,  0,                    -0.1,              1],
+    ['-0'                   ,  0,                     0,                0],
+    ;
+}
+
+# vim: expandtab shiftwidth=4
diff --git a/spectest.data b/spectest.data
index 1fa5ed5b7d..29652ea7ff 100644
--- a/spectest.data
+++ b/spectest.data
@@ -1200,6 +1200,7 @@ S32-num/fatrat.t
 S32-num/int.t
 S32-num/is-prime.t
 S32-num/log.t
+S32-num/modf.t
 S32-num/narrow.t
 S32-num/negative-zero.t
 S32-num/pi.t