diff --git l10n_community/ajax.inc l10n_community/ajax.inc
index 890e4ac..d14b82b 100644
--- l10n_community/ajax.inc
+++ l10n_community/ajax.inc
@@ -56,6 +56,9 @@ function l10n_community_string_details($langcode = NULL, $sid = 0) {
 }
 
 function l10n_community_string_suggestions($langcode = NULL, $sid = 0) {
+  // Look for an existing active translation, if any.
+  $existing_string = db_fetch_object(db_query("SELECT sid, tid, translation FROM {l10n_community_translation} WHERE sid = %d AND language = '%s' AND is_suggestion = 0 AND is_active = 1", $sid, $langcode));
+
   // Existing, "unresolved" suggestions.
   $suggestions = array();
   $result = db_query("SELECT t.tid, t.sid, t.translation, t.uid_entered, t.time_entered, u.name FROM {l10n_community_translation} t LEFT JOIN {users} u ON u.uid = t.uid_entered WHERE t.language = '%s' AND t.sid = %d AND t.is_active = 1 AND t.is_suggestion = 1 ORDER BY t.time_entered", $langcode, $sid);
@@ -71,7 +74,27 @@ function l10n_community_string_suggestions($langcode = NULL, $sid = 0) {
     // Since we no longer store an additional copy of strings in JS
     // and the null byte is stripped out by browsers (?), we need to
     // use a DOM-safe delimiter: ";  "
-    $translation = strpos($suggestion->translation, "\0") ? str_replace(chr(0), ";  ", $suggestion->translation) : $suggestion->translation;
+    if (strpos($suggestion->translation, "\0")) {
+      $translation = str_replace(chr(0), ";  ", $suggestion->translation);
+    }
+    else {
+      if (isset($existing_string->translation) && drupal_strlen($existing_string->translation) > 0) {
+        include_once 'diff.inc';
+
+        $diff = worddiff(
+          check_plain($existing_string->translation),
+          check_plain($suggestion->translation)
+        );
+
+        $translation = worddiff_render($diff);
+        $translation['original'] = $suggestion->translation;
+      }
+      else {
+        $existing_string = $suggestion;
+        $translation = $suggestion->translation;
+      }
+    }
+
     $translation = l10n_community_format_text($translation, $suggestion->sid);
     $suggestions[] = $translation ."<div class='l10n-approval-buttons'>". $approve_button . $decline_button .'</div><div class="l10n-suggestion-by">'. t('by %user at %date', array('%user' => (empty($suggestion->name) ? variable_get('anonymous', t('Anonymous')) : $suggestion->name), '%date' => format_date($suggestion->time_entered))) .'</div>';
   }
diff --git l10n_community/diff.inc l10n_community/diff.inc
new file mode 100755
index 0000000..9c7aa43
--- /dev/null
+++ l10n_community/diff.inc
@@ -0,0 +1,201 @@
+<?php
+// $Id$
+
+/**
+ * @file
+ *   Visualizes differences between translations.
+ */
+
+function _worddiff_tokenize($strings) {
+  $tokens = $length = array();
+  $prefix = $suffix = '';
+
+  // Prepare the strings.
+  foreach($strings as $key => $str) {
+    // Tokenize the string
+    $tokens[$key] = preg_split('/(?<=\W)|(?=\W)/u', trim($str));
+  }
+
+  // Remove identical prefix.
+  while ($tokens[0] && $tokens[1] && reset($tokens[0]) === reset($tokens[1])) {
+    $prefix .= array_shift($tokens[0]);
+    array_shift($tokens[1]);
+  }
+
+  // Remove identical suffixes.
+  while ($tokens[0] && $tokens[1] && end($tokens[0]) === end($tokens[1])) {
+    $suffix = array_pop($tokens[0]) . $suffix;
+    array_pop($tokens[1]);
+  }
+
+  // Calculate the length of the token arrays.
+  foreach (array_keys($tokens) as $key) {
+    $length[$key] = count($tokens[$key]);
+  }
+
+  return array($tokens, $length, $prefix, $suffix);
+}
+
+function _worddiff_lcs($tokens, $length) {
+  $matrix = array();
+
+  // Calculate the matrix.
+  for ($i = 0; $i < $length[0]; $i++) {
+    for ($j = 0; $j < $length[1]; $j++) {
+      if ($tokens[0][$i] === $tokens[1][$j]) {
+        $matrix[$i][$j] = (isset($matrix[$i - 1][$j - 1]) ? $matrix[$i - 1][$j - 1] : 0) + 1;
+      }
+      else {
+        $matrix[$i][$j] = max(
+          isset($matrix[$i][$j - 1]) ? $matrix[$i][$j - 1] : 0,
+          isset($matrix[$i - 1][$j]) ? $matrix[$i - 1][$j] : 0
+        );
+      }
+    }
+  }
+
+  return $matrix;
+}
+
+function _worddiff_expand_result($str) {
+  return array(-1 => '', 0 => $str, 1 => '');
+}
+
+function _worddiff_changeset($tokens, $matrix, $length) {
+  // Setup work.
+  $index = 0;
+  $mode = NULL;
+  $result = array_map('_worddiff_expand_result', $tokens[0]);
+
+  for ($i = $length[0] - 1, $j = $length[1] - 1; $i >= 0; $i--, $j--) {
+    if ($j < 0 || $tokens[0][$i] !== $tokens[1][$j]) {
+      if ($j < 0 || ($j > 0 && ($matrix[$i][$j - 1] < $matrix[$i - 1][$j]))) {
+        // Delete substring.
+        if ($mode === -1) {
+          // If we're already in delete mode, prepend it to the previous delete.
+          $result[$index][-1] = $result[$i][0] . $result[$index][-1];
+          unset($result[$i]);
+        }
+        else {
+          // Otherwise, move this token's string to the delete part.
+          $result[$i][-1] = $result[$i][0];
+          $result[$i][0] = '';
+          $index = $i;
+          $mode = -1;
+        }
+        $j++;
+      }
+      else {
+        // Add a new substring to the last token.
+        $result[$i][1] = $tokens[1][$j] . $result[$i][1];
+        $i++;
+        $mode = 1;
+      }
+    }
+    else {
+      // Preserve substring.
+      if ($mode === 0) {
+        // Merge this token into the previous one to minimize the amount of tokens.
+        $result[$index][0] = $result[$i][0] . $result[$index][0];
+        unset($result[$i]);
+      }
+      else {
+        $index = $i;
+        $mode = 0;
+      }
+    }
+  }
+
+  return array_values($result);
+}
+
+function _worddiff_find_prefix($a, $b) {
+  $prefix = '';
+
+  while (drupal_strlen($a) && drupal_strlen($b) && drupal_substr($a, 0, 1) === drupal_substr($b, 0, 1)) {
+    $prefix .= drupal_substr($a, 0, 1);
+    $a = drupal_substr($a, 1);
+    $b = drupal_substr($b, 1);
+  }
+
+  return array($prefix, $a, $b);
+}
+
+function _worddiff_aggregate($diff) {
+  for ($i = 0; isset($diff[$i]); $i++) {
+    if ($diff[$i][-1] == '' && drupal_strlen($diff[$i][0]) < 3 && $diff[$i][1] == '' &&
+        isset($diff[$i - 1]) && $diff[$i - 1][0] == '' &&
+        isset($diff[$i + 1]) && $diff[$i + 1][0] == '') {
+      $diff[$i + 1] = array(
+        -1 => $diff[$i - 1][-1] . $diff[$i][0] . $diff[$i + 1][-1],
+        0 => '',
+        1 => $diff[$i - 1][1] . $diff[$i][0] . $diff[$i + 1][1],
+      );
+
+      unset($diff[$i - 1], $diff[$i]);
+    }
+  }
+
+  // Find common prefixes.
+  $diff = array_values($diff);
+  for ($i = 0; isset($diff[$i]); $i++) {
+    if ($diff[$i][0] == '' && isset($diff[$i - 1]) && $diff[$i - 1][1] == '') {
+      list($prefix, $diff[$i][-1], $diff[$i][1]) = _worddiff_find_prefix($diff[$i][-1], $diff[$i][1]);
+      $diff[$i - 1][0] .= $prefix;
+    }
+  }
+
+  return $diff;
+}
+
+function worddiff($old, $new) {
+  // Preparation stuff.
+  list($tokens, $length, $prefix, $suffix) = _worddiff_tokenize(array($old, $new));
+  $matrix = _worddiff_lcs($tokens, $length);
+
+  // Generate the changeset.
+  $result = _worddiff_changeset($tokens, $matrix, $length);
+
+  // Add the previously trimmed prefix and suffix.
+  array_unshift($result, array(-1 => '', 0 => $prefix, 1 => ''));
+  array_push($result, array(-1 => '', 0 => $suffix, 1 => ''));
+
+  // Merge small changes together.
+  $result = _worddiff_aggregate($result);
+
+  return $result;
+}
+
+function _worddiff_render_deletions($changes) {
+  $value = '';
+
+  if (drupal_strlen($changes[-1])) {
+    $value .= '<del>'. $changes[-1] .'</del>';
+  }
+  if (drupal_strlen($changes[0])) {
+    $value .= $changes[0];
+  }
+
+  return $value;
+}
+
+function _worddiff_render_insertions($changes) {
+  $value = '';
+
+  if (drupal_strlen($changes[0])) {
+    $value .= $changes[0];
+  }
+  if (drupal_strlen($changes[1])) {
+    $value .= '<ins>'. $changes[1] .'</ins>';
+  }
+
+  return $value;
+}
+
+function worddiff_render($result) {
+  return array(
+    'old' => implode('', array_map('_worddiff_render_deletions', $result)),
+    'new' => implode('', array_map('_worddiff_render_insertions', $result))
+  );
+}
+
diff --git l10n_community/l10n_community.css l10n_community/l10n_community.css
index fadadbe..63b3400 100644
--- l10n_community/l10n_community.css
+++ l10n_community/l10n_community.css
@@ -61,6 +61,7 @@ div.admin form {
 
 /* Emphasized replaceable parts of translatables */
 em.l10n-community-marker {
+  font-style: normal;
   font-weight: bold;
 }
 
@@ -195,6 +196,24 @@ ul.l10n-community-strings {
   margin: 0;
 }
 
+ul.l10n-community-strings code.removed {
+  background: #FFD8D8;
+}
+
+ul.l10n-community-strings code.removed del {
+  background: #FF8888;
+  text-decoration: none;
+}
+
+ul.l10n-community-strings code.inserted {
+  background: #DDF8CC;
+}
+
+ul.l10n-community-strings code.inserted ins {
+  background: #99FF99;
+  text-decoration: none;
+}
+
 ul.l10n-community-strings li {
   padding-left: 20px;
 }
diff --git l10n_community/l10n_community.module l10n_community/l10n_community.module
index 0dc671d..5425606 100644
--- l10n_community/l10n_community.module
+++ l10n_community/l10n_community.module
@@ -1329,6 +1329,11 @@ function theme_l10n_community_copy_button() {
 function l10n_community_format_text($string, $sid = NULL, $delta = NULL) {
   static $path = NULL, $title = NULL;
 
+  if (is_array($string)) {
+    $diff = $string;
+    $string = $string['original'];
+  }
+
   if (!isset($path)) {
     $path = base_path() . drupal_get_path('module', 'l10n_community');
     $title = t('line break');
@@ -1336,12 +1341,20 @@ function l10n_community_format_text($string, $sid = NULL, $delta = NULL) {
 
   $original = check_plain($string);
 
+  if (isset($diff)) {
+    $string = '<code class="removed">' . $diff['old'] . '</code><br /><code class="inserted">' . $diff['new'] . '</code>';
+  }
+  else {
+    $string = $original;
+  }
+
   // Replace all newline chars in the string with an indicator image.
   $string = str_replace(
     array("\n", "\\\\n"),
     '<img src="'. $path .'/images/newline.png" alt="'. $title .'" title="'. $title .'" /><br />',
-    check_plain($string)
+    $string
   );
+
   // Make all %, ! and @ marked pladeholders emphasized.
   $string = preg_replace(
     '~((%|!|@)[0-9a-zA-Z_-]+)~',