diff --git a/Cargo.toml b/Cargo.toml
index 7bd41fbce..3e3ad4452 100644
--- a/Cargo.toml
+++ b/Cargo.toml
@@ -36,6 +36,10 @@ if_chain = "1.0.1"
 
 [dev-dependencies]
 annotate-snippets = { version = "0.9.0", features = ["color"] }
+# TODO(@magurotuna): This branch contains a patch that fixes how double-width
+# characters in diagnostic annotations are displayed. Once it lands, we'll
+# upgrade it.
+# annotate-snippets = { git = "https://github.com/magurotuna/annotate-snippets-rs", branch = "check-display-width", features = ["color"] }
 ansi_term = "0.12.1"
 atty = "0.2.14"
 clap = "2.33.3"
diff --git a/docs/rules/prefer_ascii.md b/docs/rules/prefer_ascii.md
new file mode 100644
index 000000000..61ea95188
--- /dev/null
+++ b/docs/rules/prefer_ascii.md
@@ -0,0 +1,47 @@
+Ensures that the code is fully written in ASCII characters.
+
+V8, the JavaScript engine Deno relies on, provides a method that strings get
+populated outside V8's heap. In particular, if they are composed of one-byte
+characters only, V8 can handle them much more efficiently through
+[`v8::String::ExternalOneByteStringResource`]. In order to leverage this V8
+feature in the internal of Deno, this rule checks if all characters in the code
+are ASCII.
+
+[`v8::String::ExternalOneByteStringResource`]: https://v8.github.io/api/head/classv8_1_1String_1_1ExternalOneByteStringResource.html
+
+That said, you can also make use of this lint rule for something other than
+Deno's internal JavaScript code. If you want to make sure your codebase is made
+up of ASCII characters only (e.g. want to disallow non-ASCII identifiers) for
+some reasons, then this rule will be helpful.
+
+### Invalid:
+
+```typescript
+const π = Math.PI;
+
+// string literals are also checked
+const ninja = "🥷";
+
+function こんにちは(名前: string) {
+  console.log(`こんにちは、${名前}さん`);
+}
+
+// “comments” are also checked
+// ^        ^
+// |        U+201D
+// U+201C
+```
+
+### Valid:
+
+```typescript
+const pi = Math.PI;
+
+const ninja = "ninja";
+
+function hello(name: string) {
+  console.log(`Hello, ${name}`);
+}
+
+// "comments" are also checked
+```
diff --git a/examples/dlint/diagnostics.rs b/examples/dlint/diagnostics.rs
new file mode 100644
index 000000000..0c6fbb30b
--- /dev/null
+++ b/examples/dlint/diagnostics.rs
@@ -0,0 +1,201 @@
+// Copyright 2020-2021 the Deno authors. All rights reserved. MIT license.
+use annotate_snippets::display_list;
+use annotate_snippets::snippet;
+use ast_view::SourceFile;
+use ast_view::SourceFileTextInfo;
+use deno_lint::diagnostic::LintDiagnostic;
+use deno_lint::diagnostic::Range;
+
+pub fn display_diagnostics(
+  diagnostics: &[LintDiagnostic],
+  source_file: &SourceFileTextInfo,
+) {
+  for diagnostic in diagnostics {
+    let (slice_source, char_range) =
+      get_slice_source_and_range(source_file, &diagnostic.range);
+    let footer = if let Some(hint) = &diagnostic.hint {
+      vec![snippet::Annotation {
+        label: Some(hint),
+        id: None,
+        annotation_type: snippet::AnnotationType::Help,
+      }]
+    } else {
+      vec![]
+    };
+
+    let snippet = snippet::Snippet {
+      title: Some(snippet::Annotation {
+        label: Some(&diagnostic.message),
+        id: Some(&diagnostic.code),
+        annotation_type: snippet::AnnotationType::Error,
+      }),
+      footer,
+      slices: vec![snippet::Slice {
+        source: slice_source,
+        line_start: diagnostic.range.start.line_index + 1, // make 1-indexed
+        origin: Some(&diagnostic.filename),
+        fold: false,
+        annotations: vec![snippet::SourceAnnotation {
+          range: char_range.as_tuple(),
+          label: "",
+          annotation_type: snippet::AnnotationType::Error,
+        }],
+      }],
+      opt: display_list::FormatOptions {
+        color: true,
+        anonymized_line_numbers: false,
+        margin: None,
+      },
+    };
+    let display_list = display_list::DisplayList::from(snippet);
+    eprintln!("{}", display_list);
+  }
+}
+
+#[derive(Debug, PartialEq, Eq)]
+struct CharRange {
+  /// 0-indexed number that represents what index this range starts at in the
+  /// snippet.
+  /// Counted on a character basis, not UTF-8 bytes.
+  start_index: usize,
+
+  /// 0-indexed number that represents what index this range ends at in the
+  /// snippet.
+  /// Counted on a character basis, not UTF-8 bytes.
+  end_index: usize,
+}
+
+impl CharRange {
+  fn as_tuple(&self) -> (usize, usize) {
+    (self.start_index, self.end_index)
+  }
+}
+
+// Return slice of source code covered by diagnostic
+// and adjusted range of diagnostic (ie. original range - start line
+// of sliced source code).
+fn get_slice_source_and_range<'a>(
+  source_file: &'a SourceFileTextInfo,
+  range: &Range,
+) -> (&'a str, CharRange) {
+  let first_line_start =
+    source_file.line_start(range.start.line_index).0 as usize;
+  let last_line_end = source_file.line_end(range.end.line_index).0 as usize;
+  let text = source_file.text();
+  let start_index =
+    text[first_line_start..range.start.byte_pos].chars().count();
+  let end_index = text[first_line_start..range.end.byte_pos].chars().count();
+  let slice_str = &text[first_line_start..last_line_end];
+  (
+    slice_str,
+    CharRange {
+      start_index,
+      end_index,
+    },
+  )
+}
+
+#[cfg(test)]
+mod tests {
+  use super::*;
+  use deno_lint::diagnostic::{Position, Range};
+  use swc_common::BytePos;
+
+  fn into_text_info(source_code: impl Into<String>) -> SourceFileTextInfo {
+    SourceFileTextInfo::new(BytePos(0), source_code.into())
+  }
+
+  fn position(byte: u32, info: &SourceFileTextInfo) -> Position {
+    let b = BytePos(byte);
+    Position::new(b, info.line_and_column_index(b))
+  }
+
+  #[test]
+  fn slice_range_a() {
+    let text_info = into_text_info("const a = 42;");
+    // 'a'
+    let range = Range {
+      start: position(6, &text_info),
+      end: position(7, &text_info),
+    };
+
+    let (slice, char_range) = get_slice_source_and_range(&text_info, &range);
+    assert_eq!(slice, "const a = 42;");
+    assert_eq!(
+      char_range,
+      CharRange {
+        start_index: 6,
+        end_index: 7,
+      }
+    );
+  }
+
+  #[test]
+  fn slice_range_あ() {
+    let text_info = into_text_info("const あ = 42;");
+    // 'あ', which takes up three bytes
+    let range = Range {
+      start: position(6, &text_info),
+      end: position(9, &text_info),
+    };
+
+    let (slice, char_range) = get_slice_source_and_range(&text_info, &range);
+    assert_eq!(slice, "const あ = 42;");
+    assert_eq!(
+      char_range,
+      CharRange {
+        start_index: 6,
+        end_index: 7,
+      }
+    );
+  }
+
+  #[test]
+  fn slice_range_あい() {
+    let text_info = into_text_info("const あい = 42;");
+    // 'い', which takes up three bytes
+    let range = Range {
+      start: position(9, &text_info),
+      end: position(12, &text_info),
+    };
+
+    let (slice, char_range) = get_slice_source_and_range(&text_info, &range);
+    assert_eq!(slice, "const あい = 42;");
+    assert_eq!(
+      char_range,
+      CharRange {
+        start_index: 7,
+        end_index: 8,
+      }
+    );
+  }
+
+  #[test]
+  fn slice_range_across_lines() {
+    let src = r#"
+const a = `あいうえお
+かきくけこ`;
+const b = 42;
+"#;
+    let text_info = into_text_info(src);
+    // "えお\nかきく"
+    let range = Range {
+      start: position(21, &text_info),
+      end: position(37, &text_info),
+    };
+
+    let (slice, char_range) = get_slice_source_and_range(&text_info, &range);
+    assert_eq!(
+      slice,
+      r#"const a = `あいうえお
+かきくけこ`;"#
+    );
+    assert_eq!(
+      char_range,
+      CharRange {
+        start_index: 14,
+        end_index: 20,
+      }
+    );
+  }
+}
diff --git a/examples/dlint/main.rs b/examples/dlint/main.rs
index 9a7181840..320e12822 100644
--- a/examples/dlint/main.rs
+++ b/examples/dlint/main.rs
@@ -1,9 +1,6 @@
 // Copyright 2020-2021 the Deno authors. All rights reserved. MIT license.
-use annotate_snippets::display_list;
-use annotate_snippets::snippet;
 use anyhow::bail;
 use anyhow::Error as AnyError;
-use ast_view::SourceFile;
 use ast_view::SourceFileTextInfo;
 use clap::App;
 use clap::AppSettings;
@@ -11,7 +8,6 @@ use clap::Arg;
 use clap::SubCommand;
 use deno_lint::ast_parser::{get_default_es_config, get_default_ts_config};
 use deno_lint::diagnostic::LintDiagnostic;
-use deno_lint::diagnostic::Range;
 use deno_lint::linter::LinterBuilder;
 use deno_lint::rules::{get_all_rules, get_recommended_rules};
 use log::debug;
@@ -24,6 +20,7 @@ use swc_ecmascript::parser::{EsConfig, Syntax, TsConfig};
 
 mod color;
 mod config;
+mod diagnostics;
 mod js;
 mod lexer;
 mod rules;
@@ -69,69 +66,6 @@ fn create_cli_app<'a, 'b>() -> App<'a, 'b> {
     )
 }
 
-// Return slice of source code covered by diagnostic
-// and adjusted range of diagnostic (ie. original range - start line
-// of sliced source code).
-fn get_slice_source_and_range<'a>(
-  source_file: &'a SourceFileTextInfo,
-  range: &Range,
-) -> (&'a str, (usize, usize)) {
-  let first_line_start =
-    source_file.line_start(range.start.line_index).0 as usize;
-  let last_line_end = source_file.line_end(range.end.line_index).0 as usize;
-  let adjusted_start = range.start.byte_pos - first_line_start;
-  let adjusted_end = range.end.byte_pos - first_line_start;
-  let adjusted_range = (adjusted_start, adjusted_end);
-  let slice_str = &source_file.text()[first_line_start..last_line_end];
-  (slice_str, adjusted_range)
-}
-
-fn display_diagnostics(
-  diagnostics: &[LintDiagnostic],
-  source_file: &SourceFileTextInfo,
-) {
-  for diagnostic in diagnostics {
-    let (slice_source, range) =
-      get_slice_source_and_range(source_file, &diagnostic.range);
-    let footer = if let Some(hint) = &diagnostic.hint {
-      vec![snippet::Annotation {
-        label: Some(hint),
-        id: None,
-        annotation_type: snippet::AnnotationType::Help,
-      }]
-    } else {
-      vec![]
-    };
-
-    let snippet = snippet::Snippet {
-      title: Some(snippet::Annotation {
-        label: Some(&diagnostic.message),
-        id: Some(&diagnostic.code),
-        annotation_type: snippet::AnnotationType::Error,
-      }),
-      footer,
-      slices: vec![snippet::Slice {
-        source: slice_source,
-        line_start: diagnostic.range.start.line_index + 1, // make 1-indexed
-        origin: Some(&diagnostic.filename),
-        fold: false,
-        annotations: vec![snippet::SourceAnnotation {
-          range,
-          label: "",
-          annotation_type: snippet::AnnotationType::Error,
-        }],
-      }],
-      opt: display_list::FormatOptions {
-        color: true,
-        anonymized_line_numbers: false,
-        margin: None,
-      },
-    };
-    let display_list = display_list::DisplayList::from(snippet);
-    eprintln!("{}", display_list);
-  }
-}
-
 fn run_linter(
   paths: Vec<String>,
   filter_rule_name: Option<&str>,
@@ -204,7 +138,7 @@ fn run_linter(
     })?;
 
   for d in file_diagnostics.lock().unwrap().values() {
-    display_diagnostics(&d.diagnostics, &d.source_file);
+    diagnostics::display_diagnostics(&d.diagnostics, &d.source_file);
   }
 
   let err_count = error_counts.load(Ordering::Relaxed);
diff --git a/src/rules.rs b/src/rules.rs
index 345b1efe6..d4bd5ed1c 100644
--- a/src/rules.rs
+++ b/src/rules.rs
@@ -83,6 +83,7 @@ pub mod no_var;
 pub mod no_window_prefix;
 pub mod no_with;
 pub mod prefer_as_const;
+pub mod prefer_ascii;
 pub mod prefer_const;
 pub mod prefer_namespace_keyword;
 pub mod prefer_primordials;
@@ -219,6 +220,7 @@ pub fn get_all_rules() -> Vec<Box<dyn LintRule>> {
     no_window_prefix::NoWindowPrefix::new(),
     no_with::NoWith::new(),
     prefer_as_const::PreferAsConst::new(),
+    prefer_ascii::PreferAscii::new(),
     prefer_const::PreferConst::new(),
     prefer_namespace_keyword::PreferNamespaceKeyword::new(),
     prefer_primordials::PreferPrimordials::new(),
diff --git a/src/rules/prefer_ascii.rs b/src/rules/prefer_ascii.rs
new file mode 100644
index 000000000..3930c3cba
--- /dev/null
+++ b/src/rules/prefer_ascii.rs
@@ -0,0 +1,174 @@
+// Copyright 2020-2021 the Deno authors. All rights reserved. MIT license.
+use super::{Context, LintRule};
+use crate::{Program, ProgramRef};
+use swc_common::{BytePos, Span};
+
+pub struct PreferAscii;
+
+const CODE: &str = "prefer-ascii";
+const MESSAGE: &str = "Non-ASCII characters are not allowed";
+
+fn hint(c: char) -> String {
+  format!("`{}` is not an ASCII. Consider replacing it", c)
+}
+
+impl LintRule for PreferAscii {
+  fn new() -> Box<Self> {
+    Box::new(PreferAscii)
+  }
+
+  fn code(&self) -> &'static str {
+    CODE
+  }
+
+  fn lint_program(&self, _context: &mut Context, _program: ProgramRef<'_>) {
+    unreachable!();
+  }
+
+  fn lint_program_with_ast_view(
+    &self,
+    context: &mut Context,
+    _program: Program<'_>,
+  ) {
+    let mut not_asciis = Vec::new();
+
+    let mut src_chars = context.source_file().text().char_indices().peekable();
+    while let Some((i, c)) = src_chars.next() {
+      if let Some(&(pi, _)) = src_chars.peek() {
+        if (pi > i + 1) || !c.is_ascii() {
+          let span = Span::new(
+            BytePos(i as u32),
+            BytePos(pi as u32),
+            Default::default(),
+          );
+          not_asciis.push((c, span));
+        }
+      }
+    }
+
+    for (c, span) in not_asciis {
+      context.add_diagnostic_with_hint(span, CODE, MESSAGE, hint(c));
+    }
+  }
+
+  #[cfg(feature = "docs")]
+  fn docs(&self) -> &'static str {
+    include_str!("../../docs/rules/prefer_ascii.md")
+  }
+}
+
+#[cfg(test)]
+mod tests {
+  use super::*;
+
+  #[test]
+  fn prefer_ascii_valid() {
+    assert_lint_ok! {
+      PreferAscii,
+      r#"const pi = Math.PI;"#,
+      r#"const ninja = "ninja";"#,
+      r#"
+function hello(name: string) {
+  console.log(`Hello, ${name}`);
+}
+      "#,
+      r#"// "comments" are also checked"#,
+      r#"/* "comments" are also checked */"#,
+    };
+  }
+
+  #[test]
+  fn prefer_ascii_invalid() {
+    assert_lint_err! {
+      PreferAscii,
+      r#"const π = Math.PI;"#: [
+        {
+          line: 1,
+          col: 6,
+          message: MESSAGE,
+          hint: hint('π'),
+        },
+      ],
+      r#"const ninja = "🥷";"#: [
+        {
+          line: 1,
+          col: 15,
+          message: MESSAGE,
+          hint: hint('🥷'),
+        },
+      ],
+      r#"function こんにちは(名前: string) {}"#: [
+        {
+          line: 1,
+          col: 9,
+          message: MESSAGE,
+          hint: hint('こ'),
+        },
+        {
+          line: 1,
+          col: 10,
+          message: MESSAGE,
+          hint: hint('ん'),
+        },
+        {
+          line: 1,
+          col: 11,
+          message: MESSAGE,
+          hint: hint('に'),
+        },
+        {
+          line: 1,
+          col: 12,
+          message: MESSAGE,
+          hint: hint('ち'),
+        },
+        {
+          line: 1,
+          col: 13,
+          message: MESSAGE,
+          hint: hint('は'),
+        },
+        {
+          line: 1,
+          col: 15,
+          message: MESSAGE,
+          hint: hint('名'),
+        },
+        {
+          line: 1,
+          col: 16,
+          message: MESSAGE,
+          hint: hint('前'),
+        },
+      ],
+      r#"// “comments” are also checked"#: [
+        {
+          line: 1,
+          col: 3,
+          message: MESSAGE,
+          hint: hint('“'),
+        },
+        {
+          line: 1,
+          col: 12,
+          message: MESSAGE,
+          hint: hint('”'),
+        },
+      ],
+      r#"/* “comments” are also checked */"#: [
+        {
+          line: 1,
+          col: 3,
+          message: MESSAGE,
+          hint: hint('“'),
+        },
+        {
+          line: 1,
+          col: 12,
+          message: MESSAGE,
+          hint: hint('”'),
+        },
+      ],
+    };
+  }
+}