diff --git a/CHANGELOG.md b/CHANGELOG.md
index 395848843..c05373b6d 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -1,8 +1,6 @@
 # Changelog
 
 # Todo
-- jslint - add numeric-separators support
-- jslint - move regexp-literals to module-level so they are explicitly cached, to improve performance
 - cli - remove cli-option `--mode-vim-plugin`
 - coverage - add macros `/*coverage-disable*/` and `/*coverage-enable*/`.
 - coverage - support globbing `*` in cli-options `--exclude` and `--include`
@@ -17,6 +15,8 @@
 - perf - improve performance by hoisting inlined regexps out of loops and subfunctions
 
 # v2022.3.1-beta
+- jslint - add numeric-separator support
+- jslint - move regexp-literals to module-level so they are explicitly cached, to improve performance
 - ci - add check for package.json.fileCount
 
 # v2022.2.20
diff --git a/jslint.mjs b/jslint.mjs
index 324089fc4..0701ee422 100644
--- a/jslint.mjs
+++ b/jslint.mjs
@@ -172,6 +172,91 @@ let jslint_export;                      // The jslint object to be exported.
 let jslint_fudge = 1;                   // Fudge starting line and starting
                                         // ... column to 1.
 let jslint_import_meta_url = "";        // import.meta.url used by cli.
+let jslint_rgx_cap = (
+    /^[A-Z]/
+);
+let jslint_rgx_crlf = (
+    /\n|\r\n?/
+);
+let jslint_rgx_digits_bits = (
+    /^[01_]*/
+);
+let jslint_rgx_digits_decimals = (
+    /^[0-9_]*/
+);
+let jslint_rgx_digits_hexs = (
+    /^[0-9A-F_]*/i
+);
+let jslint_rgx_digits_octals = (
+    /^[0-7_]*/
+);
+let jslint_rgx_directive = (
+    /^(jslint|property|global)\s+(.*)$/
+);
+let jslint_rgx_directive_part = (
+    /([a-zA-Z$_][a-zA-Z0-9$_]*)(?::\s*(true|false))?,?\s*|$/g
+);
+let jslint_rgx_identifier = (
+    /^([a-zA-Z_$][a-zA-Z0-9_$]*)$/
+);
+let jslint_rgx_json_number = (
+
+// https://datatracker.ietf.org/doc/html/rfc7159#section-6
+// number = [ minus ] int [ frac ] [ exp ]
+
+    /^-?(?:0|[1-9]\d*)(?:\.\d*)?(?:[eE][\-+]?\d+)?$/
+);
+let jslint_rgx_mega = (
+
+// Vim-hack - vim-editor has trouble parsing naked '`' in regexp
+
+    /[\u0060\\]|\$\{/
+);
+let jslint_rgx_module = (
+    /^[a-zA-Z0-9_$:.@\-\/]+$/
+);
+let jslint_rgx_numeric_separator_illegal = (
+    /__|_$|_n$/m
+);
+let jslint_rgx_slash_star_or_slash = (
+    /\/\*|\/$/
+);
+let jslint_rgx_tab = (
+    /\t/g
+);
+let jslint_rgx_todo = (
+    /\b(?:todo|TO\s?DO|HACK)\b/
+);
+let jslint_rgx_token = new RegExp(
+    "^("
+    + "(\\s+)"
+    + "|([a-zA-Z_$][a-zA-Z0-9_$]*)"
+    + "|[(){}\\[\\],:;'\"~\\`]"
+    + "|\\?[?.]?"
+    + "|=(?:==?|>)?"
+    + "|\\.+"
+    + "|\\*[*\\/=]?"
+    + "|\\/[*\\/]?"
+    + "|\\+[=+]?"
+    + "|-[=\\-]?"
+    + "|[\\^%]=?"
+    + "|&[&=]?"
+    + "|\\"
+    + "|[|=]?"
+    + "|>{1,3}=?"
+    + "|<<?=?"
+    + "|!(?:!|==?)?"
+
+// PR-351 - Add BigInt support.
+// PR-390 - Add numeric-separator support.
+
+    + "|((?:0_?|[1-9][0-9_]*)n?)"
+    + ")"
+    + "(.*)$"
+);
+let jslint_rgx_weird_property = (
+    /^_|\$|Sync$|_$/m
+);
 let jstestCountFailed = 0;
 let jstestCountTotal = 0;
 let jstestItCount = 0;
@@ -307,10 +392,7 @@ function jslint(
     let import_list = [];       // The array collecting all import-from strings.
     let line_list = String(     // The array containing source lines.
         "\n" + source
-    ).split(
-        // rx_crlf
-        /\n|\r\n?/
-    ).map(function (line_source) {
+    ).split(jslint_rgx_crlf).map(function (line_source) {
         return {
             line_source
         };
@@ -687,6 +769,12 @@ function jslint(
 //         case "function_in_loop":
 //             mm = `Don't create functions within a loop.`;
 //             break;
+
+// PR-390 - Add numeric-separator check.
+
+        case "illegal_num_separator":
+            mm = `Illegal numeric separator '_' at column ${column}.`;
+            break;
         case "infix_in":
             mm = (
                 `Unexpected 'in'. Compare with undefined,`
@@ -1841,6 +1929,8 @@ function jslint_phase2_lex(state) {
     let line_mega;              // The starting line of megastring.
     let line_source = "";       // The remaining line source string.
     let line_whole = "";        // The whole line source string.
+    let mode_digits_empty_string = 1;
+    let mode_digits_numeric_separator = 2;
     let mode_directive = true;  // true if directives are still allowed.
     let mode_mega = false;      // true if currently parsing a megastring
                                 // ... literal.
@@ -1848,33 +1938,7 @@ function jslint_phase2_lex(state) {
                                 // ... this line.
     let paren_backtrack_list = [];      // List of most recent "(" tokens at any
                                         // ... paren-depth.
-    let paren_depth = 0;                // Keeps track of current paren-depth.
-    let rx_token = new RegExp(
-        "^("
-        + "(\\s+)"
-        + "|([a-zA-Z_$][a-zA-Z0-9_$]*)"
-        + "|[(){}\\[\\],:;'\"~\\`]"
-        + "|\\?[?.]?"
-        + "|=(?:==?|>)?"
-        + "|\\.+"
-        + "|\\*[*\\/=]?"
-        + "|\\/[*\\/]?"
-        + "|\\+[=+]?"
-        + "|-[=\\-]?"
-        + "|[\\^%]=?"
-        + "|&[&=]?"
-        + "|\\"
-        + "|[|=]?"
-        + "|>{1,3}=?"
-        + "|<<?=?"
-        + "|!(?:!|==?)?"
-
-// PR-351 - Add BigInt support.
-
-        + "|(0n?|[1-9][0-9]*n?)"
-        + ")"
-        + "(.*)$"
-    );
+    let paren_depth = 0;        // Keeps track of current paren-depth.
     let snippet = "";           // A piece of string.
     let token_1;                // The first token.
     let token_prv = token_global;       // The previous token including
@@ -1956,7 +2020,7 @@ function jslint_phase2_lex(state) {
 
                     warn_at("unexpected_a", line, column, char);
                 }
-                if (read_digits("x") > 5) {
+                if (read_digits("x", undefined) > 5) {
 
 // test_cause:
 // ["\"\\u{123456}\"", "char_after_escape", "too_many_digits", "", 11]
@@ -1973,7 +2037,7 @@ function jslint_phase2_lex(state) {
                 return char_after();
             }
             char_before();
-            if (read_digits("x", true) < 4) {
+            if (read_digits("x", mode_digits_empty_string) < 4) {
 
 // test_cause:
 // ["\"\\u0\"", "char_after_escape", "expected_four_digits", "", 5]
@@ -2008,6 +2072,26 @@ function jslint_phase2_lex(state) {
         return char;
     }
 
+    function check_numeric_separator(digits, column) {
+
+// This function will check for illegal numeric-separator in <digits>.
+
+        digits.replace((
+            jslint_rgx_numeric_separator_illegal
+        ), function (ignore, ii) {
+
+// test_cause:
+// ["0x0_0_;", "check_numeric_separator", "illegal_num_separator", "", 6]
+// ["0x0_0__0;", "check_numeric_separator", "illegal_num_separator", "", 6]
+// ["aa=1_2_;", "check_numeric_separator", "illegal_num_separator", "", 7]
+// ["aa=1_2__3;", "check_numeric_separator", "illegal_num_separator", "", 7]
+// ["aa=1_2_n;", "check_numeric_separator", "illegal_num_separator", "", 7]
+
+            warn_at("illegal_num_separator", line, column + ii + 1);
+            return "";
+        });
+    }
+
     function lex_comment() {
         let body;
         let ii = 0;
@@ -2046,12 +2130,12 @@ function jslint_phase2_lex(state) {
 // Lex/loop through each line until "*/".
 
             while (true) {
-                // rx_star_slash
+                // jslint_rgx_star_slash
                 ii = line_source.indexOf("*/");
                 if (ii >= 0) {
                     break;
                 }
-                // rx_slash_star
+                // jslint_rgx_slash_star
                 ii = line_source.indexOf("/*");
                 if (ii >= 0) {
 
@@ -2071,8 +2155,7 @@ function jslint_phase2_lex(state) {
                 }
             }
             jj = line_source.slice(0, ii).search(
-                // rx_slash_star_or_slash
-                /\/\*|\/$/
+                jslint_rgx_slash_star_or_slash
             );
             if (jj >= 0) {
 
@@ -2090,13 +2173,7 @@ function jslint_phase2_lex(state) {
 
 // Uncompleted work comment.
 
-        if (
-            !option_dict.devel
-            && (
-                // rx_todo
-                /\b(?:todo|TO\s?DO|HACK)\b/
-            ).test(snippet)
-        ) {
+        if (!option_dict.devel && jslint_rgx_todo.test(snippet)) {
 
 // test_cause:
 // ["//todo", "lex_comment", "todo_comment", "(comment)", 1] //jslint-quiet
@@ -2108,10 +2185,7 @@ function jslint_phase2_lex(state) {
 
         [
             the_comment.directive, body
-        ] = Array.from(snippet.match(
-            // rx_directive
-            /^(jslint|property|global)\s+(.*)$/
-        ) || []).slice(1);
+        ] = Array.from(snippet.match(jslint_rgx_directive) || []).slice(1);
         if (the_comment.directive === undefined) {
             return the_comment;
         }
@@ -2133,10 +2207,12 @@ function jslint_phase2_lex(state) {
 // Lex/loop through each directive in /*...*/
 
         ii = 0;
-        body.replace((
-            // rx_directive_part
-            /([a-zA-Z$_][a-zA-Z0-9$_]*)(?::\s*(true|false))?,?\s*|$/g
-        ), function (match0, key, val, jj) {
+        body.replace(jslint_rgx_directive_part, function (
+            match0,
+            key,
+            val,
+            jj
+        ) {
             if (ii !== jj) {
 
 // test_cause:
@@ -2210,13 +2286,7 @@ function jslint_phase2_lex(state) {
 // string.
 
         while (true) {
-            match = line_source.match(
-
-// Vim-hack - vim-editor has trouble parsing '`' in regexp
-
-                // rx_mega
-                /[\u0060\\]|\$\{/
-            ) || {
+            match = line_source.match(jslint_rgx_mega) || {
                 "0": "",
                 index: 0
             };
@@ -2293,12 +2363,16 @@ function jslint_phase2_lex(state) {
 
     function lex_number() {
         let prefix = snippet;
+
+// PR-390 - Add numeric-separator check.
+
+        check_numeric_separator(prefix, column - prefix.length);
         char_after();
         switch (prefix === "0" && char) {
         case "b":
         case "o":
         case "x":
-            read_digits(char);
+            read_digits(char, mode_digits_numeric_separator);
 
 // PR-351 - Ignore BigInt suffix 'n'.
 
@@ -2308,14 +2382,14 @@ function jslint_phase2_lex(state) {
             break;
         default:
             if (char === ".") {
-                read_digits("d");
+                read_digits("d", mode_digits_numeric_separator);
             }
             if (char === "E" || char === "e") {
                 char_after(char);
                 if (char !== "+" && char !== "-") {
                     char_before();
                 }
-                read_digits("d");
+                read_digits("d", mode_digits_numeric_separator);
             }
         }
 
@@ -2585,7 +2659,7 @@ function jslint_phase2_lex(state) {
                     }
                     break;
                 case "{":
-                    if (read_digits("d", true) === 0) {
+                    if (read_digits("d", mode_digits_empty_string) === 0) {
 
 // test_cause:
 // ["aa=/aa{/", "lex_regexp_group", "expected_a_before_b", ",", 8]
@@ -2598,7 +2672,7 @@ function jslint_phase2_lex(state) {
 // ["aa=/.{,/", "lex_regexp_group", "comma", "", 0]
 
                         test_cause("comma");
-                        read_digits("d", true);
+                        read_digits("d", mode_digits_empty_string);
                     }
                     if (char_after("}") === "?") {
 
@@ -2891,7 +2965,7 @@ function jslint_phase2_lex(state) {
                 }
             }
             from = column;
-            match = line_source.match(rx_token);
+            match = line_source.match(jslint_rgx_token);
 
 // match[1] token
 // match[2] whitespace
@@ -3274,41 +3348,48 @@ import moduleHttps from "https";
         return true;
     }
 
-    function read_digits(base, quiet) {
+    function read_digits(base, mode) {
         let digits = line_source.match(
             base === "b"
-            ? (
-                // rx_bits
-                /^[01]*/
-            )
+            ? jslint_rgx_digits_bits
             : base === "o"
-            ? (
-                // rx_octals
-                /^[0-7]*/
-            )
+            ? jslint_rgx_digits_octals
             : base === "x"
-            ? (
-                // rx_hexs
-                /^[0-9A-F]*/i
-            )
-            : (
-                // rx_digits
-                /^[0-9]*/
-            )
+            ? jslint_rgx_digits_hexs
+            : jslint_rgx_digits_decimals
         )[0];
-        let length = digits.length;
-        if (!quiet && length === 0) {
+        if (
+            (mode !== mode_digits_empty_string && digits.length === 0)
+            || digits[0] === "_"
+        ) {
 
 // test_cause:
 // ["0x", "read_digits", "expected_digits_after_a", "0x", 2]
+// ["0x_", "read_digits", "expected_digits_after_a", "0x", 2]
 
             warn_at("expected_digits_after_a", line, column, snippet);
         }
-        column += length;
-        line_source = line_source.slice(length);
+
+// PR-390 - Add numeric-separator check.
+
+        if (mode === mode_digits_numeric_separator) {
+            check_numeric_separator(digits, column);
+        } else if (digits.indexOf("_") >= 0) {
+
+// test_cause:
+// ["\"\\u{1_2}\"", "read_digits", "illegal_num_separator", "", 6]
+
+            warn_at(
+                "illegal_num_separator",
+                line,
+                column + digits.indexOf("_") + 1
+            );
+        }
+        column += digits.length;
+        line_source = line_source.slice(digits.length);
         snippet += digits;
         char_after();
-        return length;
+        return digits.length;
     }
 
     function read_line() {
@@ -3379,6 +3460,7 @@ import moduleHttps from "https";
             test_cause("line_disable");
             line_source = "";
         }
+        // jslint_rgx_tab
         if (line_source.indexOf("\t") >= 0) {
             if (!option_dict.white) {
 
@@ -3387,10 +3469,7 @@ import moduleHttps from "https";
 
                 warn_at("use_spaces", line, line_source.indexOf("\t") + 1);
             }
-            line_source = line_source.replace((
-                // rx_tab
-                /\t/g
-            ), " ");
+            line_source = line_source.replace(jslint_rgx_tab, " ");
         }
         if (!option_dict.white && line_source.endsWith(" ")) {
 
@@ -3564,9 +3643,6 @@ function jslint_phase3_parse(state) {
     let catchage = catch_stack[0];      // The current catch-block.
     let functionage = token_global;     // The current function.
     let mode_var;               // "var" if using var; "let" if using let.
-    let rx_identifier = (
-        /^([a-zA-Z_$][a-zA-Z0-9_$]*)$/
-    );
     let token_ii = 0;           // The number of the next token.
     let token_now = token_global;       // The current token being examined in
                                         // ... the parse.
@@ -3583,7 +3659,7 @@ function jslint_phase3_parse(state) {
             anon = token_now.id;
         } else if (
             token_now.id === "(string)"
-            && rx_identifier.test(token_now.value)
+            && jslint_rgx_identifier.test(token_now.value)
         ) {
             anon = token_now.value;
         }
@@ -4233,7 +4309,7 @@ function jslint_phase3_parse(state) {
         let the_subscript = parse_expression(0);
         if (the_subscript.id === "(string)" || the_subscript.id === "`") {
             name = survey(the_subscript);
-            if (rx_identifier.test(name)) {
+            if (jslint_rgx_identifier.test(name)) {
 
 // test_cause:
 // ["aa[`aa`]", "infix_lbracket", "subscript_a", "aa", 4]
@@ -4594,13 +4670,7 @@ function jslint_phase3_parse(state) {
         let negative;
         switch (token_nxt.id) {
         case "(number)":
-            if (!(
-
-// https://datatracker.ietf.org/doc/html/rfc7159#section-6
-// number = [ minus ] int [ frac ] [ exp ]
-
-                /^-?(?:0|[1-9]\d*?)(?:\.\d*?)?(?:[eE][+\-]?\d+?)?$/
-            ).test(token_nxt.value)) {
+            if (!jslint_rgx_json_number.test(token_nxt.value)) {
 
 // test_cause:
 // ["[-.0]", "parse_json", "unexpected_a", ".", 3]
@@ -6100,10 +6170,7 @@ function jslint_phase3_parse(state) {
         advance("from");
         advance("(string)");
         the_import.import = token_now;
-        if (!(
-            // rx_module
-            /^[a-zA-Z0-9_$:.@\-\/]+$/
-        ).test(token_now.value)) {
+        if (!jslint_rgx_module.test(token_now.value)) {
 
 // test_cause:
 // ["import aa from \"!aa\"", "stmt_import", "bad_module_name_a", "!aa", 16]
@@ -6714,13 +6781,13 @@ function jslint_phase3_parse(state) {
 
         if (id === "(string)") {
             id = name.value;
-            if (!rx_identifier.test(id)) {
+            if (!jslint_rgx_identifier.test(id)) {
                 return id;
             }
         } else if (id === "`") {
             if (name.value.length === 1) {
                 id = name.value[0].value;
-                if (!rx_identifier.test(id)) {
+                if (!jslint_rgx_identifier.test(id)) {
                     return id;
                 }
             }
@@ -6753,10 +6820,7 @@ function jslint_phase3_parse(state) {
             } else if (
                 !option_dict.name
                 && name.identifier
-                && (
-                    // rx_weird_property
-                    /^_|\$|Sync$|_$/m
-                ).test(id)
+                && jslint_rgx_weird_property.test(id)
             ) {
 
 // test_cause:
@@ -7510,10 +7574,7 @@ function jslint_phase4_walk(state) {
                 test_cause("cack");
                 cack = !cack;
             }
-            if ((
-                // rx_cap
-                /^[A-Z]/
-            ).test(left.name.id) !== cack) {
+            if (jslint_rgx_cap.test(left.name.id) !== cack) {
                 if (the_new !== undefined) {
 
 // test_cause:
diff --git a/test.mjs b/test.mjs
index f33e8d493..5beb7284d 100644
--- a/test.mjs
+++ b/test.mjs
@@ -391,6 +391,16 @@ jstestDescribe((
                 "let aa = 0o0;",
                 "let aa = 0x0;"
             ],
+
+// PR-390 - Add numeric-separator support.
+
+            numeric_separator: [
+                "let aa = 0.0_0_0;",
+                "let aa = 0b0_1111_1111n;\n",
+                "let aa = 0o0_1234_1234n;\n",
+                "let aa = 0x0_1234_1234n;\n",
+                "let aa = 1_234_234.1_234_234E1_234_234;"
+            ],
             optional_chaining: [
                 "let aa = aa?.bb?.cc;"
             ],