diff --git a/bin/generate-run-length-encoding b/bin/generate-run-length-encoding
new file mode 100755
index 0000000000..56dcca121c
--- /dev/null
+++ b/bin/generate-run-length-encoding
@@ -0,0 +1,6 @@
+#!/usr/bin/env ruby
+
+require_relative '../lib/generator'
+require_relative '../lib/run_length_encoding_cases'
+
+Generator.new('run-length-encoding', RunLengthEncodingCases).generate
diff --git a/config.json b/config.json
index b5d7cc2ad4..e0f9bc0348 100644
--- a/config.json
+++ b/config.json
@@ -19,6 +19,7 @@
     "word-count",
     "bob",
     "food-chain",
+    "run-length-encoding",
     "sieve",
     "binary",
     "accumulate",
diff --git a/exercises/run-length-encoding/.version b/exercises/run-length-encoding/.version
new file mode 100644
index 0000000000..56a6051ca2
--- /dev/null
+++ b/exercises/run-length-encoding/.version
@@ -0,0 +1 @@
+1
\ No newline at end of file
diff --git a/exercises/run-length-encoding/example.rb b/exercises/run-length-encoding/example.rb
new file mode 100644
index 0000000000..bde2e01429
--- /dev/null
+++ b/exercises/run-length-encoding/example.rb
@@ -0,0 +1,32 @@
+class RunLengthEncoding
+  VERSION = 1
+
+  def self.encode(str)
+    str.chars.chunk { |char| char }.each_with_object('') do |chunk, out|
+      out << encoded(chunk)
+    end
+  end
+
+  def self.decode(str)
+    str.scan(/(\d+)?(\D)/).each_with_object('') do |captures, out|
+      out << decoded(captures)
+    end
+  end
+
+  # private
+
+  def self.encoded(chunk)
+    char = chunk.first
+    times = chunk.last.count
+    return char if times == 1
+    "#{times}#{char}"
+  end
+  private_class_method :encoded
+
+  def self.decoded(captures)
+    times = (captures.first || 1).to_i
+    char = captures.last
+    char * times
+  end
+  private_class_method :decoded
+end
diff --git a/exercises/run-length-encoding/example.tt b/exercises/run-length-encoding/example.tt
new file mode 100644
index 0000000000..62efd5e157
--- /dev/null
+++ b/exercises/run-length-encoding/example.tt
@@ -0,0 +1,29 @@
+#!/usr/bin/env ruby
+# encoding: utf-8
+gem 'minitest', '>= 5.0.0'
+require 'minitest/autorun'
+require_relative 'run_length_encoding'
+
+# Test data version:
+# <%= sha1 %>
+class RunLengthEncodingTest < Minitest::Test<% test_cases.each do |test_case| %>
+  def <%= test_case.name %><% if test_case.skipped? %>
+    skip<% end %>
+    <%= test_case.assign_input %>
+    <%= test_case.assign_output %>
+    <%= test_case.assertion %>
+  end
+<% end %>
+  # Problems in exercism evolve over time,
+  # as we find better ways to ask questions.
+  # The version number refers to the version of the problem you solved,
+  # not your solution.
+  #
+  # Define a constant named VERSION inside of RunLengthEncoding.
+  # If you are curious, read more about constants on RubyDoc:
+  # http://ruby-doc.org/docs/ruby-doc-bundle/UsersGuide/rg/constants.html
+  def test_bookkeeping
+    skip
+    assert_equal <%= version+1 %>, RunLengthEncoding::VERSION
+  end
+end
diff --git a/exercises/run-length-encoding/run_length_encoding_test.rb b/exercises/run-length-encoding/run_length_encoding_test.rb
new file mode 100755
index 0000000000..d798dd4e24
--- /dev/null
+++ b/exercises/run-length-encoding/run_length_encoding_test.rb
@@ -0,0 +1,72 @@
+#!/usr/bin/env ruby
+# encoding: utf-8
+gem 'minitest', '>= 5.0.0'
+require 'minitest/autorun'
+require_relative 'run_length_encoding'
+
+# Test data version:
+# eb8d142 Merge pull request #220 from IanWhitney/sieve_ordering
+
+class RunLengthEncodingTest < Minitest::Test
+  def test_encode_simple
+    input = 'AABBBCCCC'
+    output = '2A3B4C'
+    assert_equal output, RunLengthEncoding.encode(input)
+  end
+
+  def test_decode_simple
+    skip
+    input = '2A3B4C'
+    output = 'AABBBCCCC'
+    assert_equal output, RunLengthEncoding.decode(input)
+  end
+
+  def test_encode_with_single_values
+    skip
+    input = 'WWWWWWWWWWWWBWWWWWWWWWWWWBBBWWWWWWWWWWWWWWWWWWWWWWWWB'
+    output = '12WB12W3B24WB'
+    assert_equal output, RunLengthEncoding.encode(input)
+  end
+
+  def test_decode_with_single_values
+    skip
+    input = '12WB12W3B24WB'
+    output = 'WWWWWWWWWWWWBWWWWWWWWWWWWBBBWWWWWWWWWWWWWWWWWWWWWWWWB'
+    assert_equal output, RunLengthEncoding.decode(input)
+  end
+
+  def test_decode_encode_combination
+    skip
+    input = 'zzz ZZ  zZ'
+    output = 'zzz ZZ  zZ'
+    assert_equal output,
+                 RunLengthEncoding.decode(RunLengthEncoding.encode(input))
+  end
+
+  def test_encode_unicode
+    skip
+    input = '⏰⚽⚽⚽⭐⭐⏰'
+    output = '⏰3⚽2⭐⏰'
+    assert_equal output, RunLengthEncoding.encode(input)
+  end
+
+  def test_decode_unicode
+    skip
+    input = '⏰3⚽2⭐⏰'
+    output = '⏰⚽⚽⚽⭐⭐⏰'
+    assert_equal output, RunLengthEncoding.decode(input)
+  end
+
+  # Problems in exercism evolve over time,
+  # as we find better ways to ask questions.
+  # The version number refers to the version of the problem you solved,
+  # not your solution.
+  #
+  # Define a constant named VERSION inside of RunLengthEncoding.
+  # If you are curious, read more about constants on RubyDoc:
+  # http://ruby-doc.org/docs/ruby-doc-bundle/UsersGuide/rg/constants.html
+  def test_bookkeeping
+    skip
+    assert_equal 1, RunLengthEncoding::VERSION
+  end
+end
diff --git a/lib/run_length_encoding_cases.rb b/lib/run_length_encoding_cases.rb
new file mode 100644
index 0000000000..32ce39d763
--- /dev/null
+++ b/lib/run_length_encoding_cases.rb
@@ -0,0 +1,41 @@
+class RunLengthEncodingCase < OpenStruct
+  def name
+    'test_%s' % cleaned_description
+  end
+
+  def assign_input
+    "input = '#{input}'"
+  end
+
+  def assign_output
+    "output = '#{expected}'"
+  end
+
+  def assertion
+    case description
+    when /decode.+encode/
+      'assert_equal output,
+                 RunLengthEncoding.decode(RunLengthEncoding.encode(input))'
+    when /encode/
+      'assert_equal output, RunLengthEncoding.encode(input)'
+    when /decode/
+      'assert_equal output, RunLengthEncoding.decode(input)'
+    end
+  end
+
+  def skipped?
+    index > 0
+  end
+
+  # internal
+
+  def cleaned_description
+    description.gsub(/\W+/, '_').squeeze('_')
+  end
+end
+
+RunLengthEncodingCases = proc do |data|
+  JSON.parse(data)['cases'].map.with_index do |row, i|
+    RunLengthEncodingCase.new(row.merge('index' => i))
+  end
+end