diff --git a/starlark/ast.h b/starlark/ast.h
index fc8b546..03cdd83 100644
--- a/starlark/ast.h
+++ b/starlark/ast.h
@@ -5,6 +5,7 @@
 #ifndef STARLARK_AST_H_
 #define STARLARK_AST_H_
 
+#include <cstdint>
 #include <memory>
 #include <optional>
 #include <string>
@@ -19,6 +20,11 @@ struct StringLiteral {
     constexpr bool operator==(StringLiteral const &) const = default;
 };
 
+struct IntLiteral {
+    std::int64_t value{};
+    constexpr bool operator==(IntLiteral const &) const = default;
+};
+
 struct Identifier {
     std::string name;
     constexpr bool operator==(Identifier const &) const = default;
@@ -28,7 +34,8 @@ struct CallExpr;
 struct DictExpr;
 struct ListExpr;
 struct ListComp;
-using Expression = std::variant<CallExpr, StringLiteral, Identifier, ListComp, ListExpr, DictExpr>;
+using Expression =
+    std::variant<CallExpr, StringLiteral, IntLiteral, Identifier, ListComp, ListExpr, DictExpr>;
 
 struct Argument;
 
diff --git a/starlark/parser.h b/starlark/parser.h
index e82e106..24c89b5 100644
--- a/starlark/parser.h
+++ b/starlark/parser.h
@@ -108,6 +108,10 @@ class Parser {
             return StringLiteral{.value = std::move(sl->value)};
         }
 
+        if (auto *il = std::get_if<token::IntLiteral>(&token)) {
+            return IntLiteral{.value = il->value};
+        }
+
         if (std::holds_alternative<token::LBracket>(token)) {
             std::vector<Expression> elements;
             while (true) {
@@ -190,8 +194,8 @@ class Parser {
                     continue;
                 }
 
-                std::cerr << "Expected ',' or ']' in list expression, got " << to_string(next_token)
-                          << ".\n";
+                std::cerr << "Expected ',' or ']' in list expression, got '"
+                          << to_string(next_token) << "'.\n";
                 return std::nullopt;
             }
 
@@ -255,8 +259,8 @@ class Parser {
                     continue;
                 }
 
-                std::cerr << "Expected ',' or '}' in dict expression, got " << to_string(next_token)
-                          << ".\n";
+                std::cerr << "Expected ',' or '}' in dict expression, got '"
+                          << to_string(next_token) << "'.\n";
                 return std::nullopt;
             }
 
@@ -311,8 +315,8 @@ class Parser {
                 auto const &token = *maybe_token;
                 if (!std::holds_alternative<token::Comma>(token) &&
                     !std::holds_alternative<token::RParen>(token)) {
-                    std::cerr << "Expected ',' or ')' in argument list, got " << to_string(token)
-                              << ".\n";
+                    std::cerr << "Expected ',' or ')' in argument list, got '" << to_string(token)
+                              << "'.\n";
                     return std::nullopt;
                 }
 
@@ -408,8 +412,8 @@ class Parser {
             }
 
             if (!std::holds_alternative<token::Comma>(*maybe_comma_or_rparen)) {
-                std::cerr << "Expected ',' or ')' in load statement, got "
-                          << to_string(*maybe_comma_or_rparen) << ".\n";
+                std::cerr << "Expected ',' or ')' in load statement, got '"
+                          << to_string(*maybe_comma_or_rparen) << "'.\n";
                 return std::nullopt;
             }
 
@@ -459,8 +463,8 @@ class Parser {
         }
 
         if (next != expected) {
-            std::cerr << "Expected " << to_string(expected) << ", got " << to_string(*next)
-                      << ".\n";
+            std::cerr << "Expected " << to_string(expected) << ", got '" << to_string(*next)
+                      << "'.\n";
             return false;
         }
 
@@ -479,8 +483,8 @@ class Parser {
             return std::move(*t);
         }
 
-        std::cerr << "Expected token of type " << typeid(T).name() << ", got " << to_string(*next)
-                  << ".\n";
+        std::cerr << "Expected token of type " << typeid(T).name() << ", got '" << to_string(*next)
+                  << "'.\n";
         return std::nullopt;
     }
 };
diff --git a/starlark/parser_test.cc b/starlark/parser_test.cc
index f91359a..0db7a51 100644
--- a/starlark/parser_test.cc
+++ b/starlark/parser_test.cc
@@ -249,6 +249,16 @@ int main() {
                 },
             },
         },
+        {
+            "42",
+            starlark::Program{
+                .statements{
+                    starlark::ExpressionStmt{
+                        .expr{starlark::IntLiteral{42}},
+                    },
+                },
+            },
+        },
     });
 
     // TODO(robinlinden): Return error codes from parser and use that here.
diff --git a/starlark/token.h b/starlark/token.h
index bf92a43..6b1c0a9 100644
--- a/starlark/token.h
+++ b/starlark/token.h
@@ -5,6 +5,7 @@
 #ifndef STARLARK_TOKEN_H_
 #define STARLARK_TOKEN_H_
 
+#include <cstdint>
 #include <format>
 #include <string>
 #include <string_view>
@@ -311,6 +312,13 @@ struct Identifier {
 
 constexpr std::string_view to_string(Identifier const &id) { return id.name; }
 
+struct IntLiteral {
+    std::int64_t value{};
+    constexpr bool operator==(IntLiteral const &) const = default;
+};
+
+inline std::string to_string(IntLiteral const &i) { return std::to_string(i.value); }
+
 struct StringLiteral {
     std::string value;
     constexpr bool operator==(StringLiteral const &) const = default;
@@ -384,6 +392,7 @@ using Token = std::variant<
     token::Lambda,
     token::Return,
     token::Identifier,
+    token::IntLiteral,
     token::StringLiteral,
     token::Eof>;
 
diff --git a/starlark/tokenizer.h b/starlark/tokenizer.h
index 7d0150b..519069a 100644
--- a/starlark/tokenizer.h
+++ b/starlark/tokenizer.h
@@ -10,10 +10,12 @@
 #include <algorithm>
 #include <array>
 #include <cassert>
+#include <charconv>
 #include <cstddef>
 #include <optional>
 #include <string>
 #include <string_view>
+#include <system_error>
 #include <utility>
 #include <variant>
 #include <vector>
@@ -43,11 +45,14 @@ class Tokenizer {
             return tokenize_identifier();
         }
 
+        if (is_digit(input_[pos_]) ||
+            (input_[pos_] == '-' && pos_ + 1 < input_.size() && is_digit(input_[pos_ + 1]))) {
+            return tokenize_number();
+        }
+
         return tokenize_punctuator();
     }
 
-    std::string_view remaining_input() const { return input_.substr(pos_); }
-
   private:
     bool is_whitespace(char c) const { return c == ' ' || c == '\t' || c == '\n' || c == '\r'; }
 
@@ -73,6 +78,30 @@ class Tokenizer {
         }
     }
 
+    std::optional<Token> tokenize_number() {
+        assert(is_digit(input_[pos_]) || input_[pos_] == '-');
+
+        std::size_t start = pos_++;
+        while (pos_ < input_.size() && is_digit(input_[pos_])) {
+            ++pos_;
+        }
+
+        // TODO(robinlinden): Support floats.
+        if (pos_ < input_.size() && (is_alpha(input_[pos_]) || input_[pos_] == '.')) {
+            return std::nullopt;
+        }
+
+        auto numstr = input_.substr(start, pos_ - start);
+
+        std::int64_t value{};
+        auto [ptr, ec] = std::from_chars(numstr.data(), numstr.data() + numstr.size(), value);
+        if (ec != std::errc{} || ptr != numstr.data() + numstr.size()) {
+            return std::nullopt;
+        }
+
+        return token::IntLiteral{value};
+    }
+
     // TODO(robinlinden): Support escapes.
     std::optional<Token> tokenize_multiline_string() {
         pos_ += 3;                // Move past the opening triple quotes
diff --git a/starlark/tokenizer_test.cc b/starlark/tokenizer_test.cc
index d176992..9a7f1a7 100644
--- a/starlark/tokenizer_test.cc
+++ b/starlark/tokenizer_test.cc
@@ -10,6 +10,8 @@
 
 #include <array>
 #include <cassert>
+#include <cstdint>
+#include <limits>
 #include <optional>
 #include <string>
 #include <string_view>
@@ -18,6 +20,7 @@
 
 int main() {
     namespace t = starlark::token;
+    using Tokens = std::vector<starlark::Token>;
 
     auto const test_cases =
         std::to_array<std::pair<std::string_view, std::optional<std::vector<starlark::Token>>>>({
@@ -36,6 +39,21 @@ int main() {
             },
             {"global", std::nullopt}, // Reserved identifier.
             {"globalist", std::vector<starlark::Token>{t::Identifier{"globalist"}}},
+            {"1234", Tokens{t::IntLiteral{1234}}},
+            {"00001234", Tokens{t::IntLiteral{1234}}},
+            {"123abc", std::nullopt},
+            {"123.123", std::nullopt}, // TODO(robinlinden): Floats.
+            {"-123", Tokens{t::IntLiteral{-123}}},
+            {
+                "9223372036854775807",
+                Tokens{t::IntLiteral{std::numeric_limits<std::int64_t>::max()}},
+            },
+            {"9223372036854775808", std::nullopt}, // Out of range. :(
+            {
+                "-9223372036854775808",
+                Tokens{t::IntLiteral{std::numeric_limits<std::int64_t>::min()}},
+            },
+            {"-9223372036854775809", std::nullopt}, // Out of range. :(
         });
 
     etest::Suite s{};