The upside down question mark alt code is Alt+0191 on Windows computers with a numeric keypad. Hold Alt, type 0191, release Alt, and the ¿ character appears. Mac and Linux users need different methods, and laptops without a dedicated keypad require workarounds explained below.

The ¿ symbol is the inverted or "turned" question mark, written U+00BF in Unicode and called "inverted question mark" in most character maps. Spanish uses it to open interrogative sentences, which lets readers know a question is coming before they reach the end. The character has existed in typewriter and computing standards for decades, so almost every keyboard layout and device has some way to produce it.

This guide covers the exact keystrokes, why they work at the character-encoding level, where they fail, and how to type the symbol on hardware that lacks a numeric pad.

How the Windows alt code mechanism works

Windows alt codes are decimal references into the system's legacy code page, not modern Unicode values. Alt+0191 works because 191 is the decimal position of ¿ in Windows-1252, the Western European code page that Windows has used since the 1990s. When you hold Alt and type digits on the numeric keypad, the keyboard driver passes that decimal value to the application, which converts it through the active code page.

You must type the leading zero for Windows-1252 characters. Alt+191 alone often fails because Windows tries to interpret it as a value in a different or stale code page. The 0191 four-digit form forces the system into the correct mapping. This is why some online lists give three-digit codes and others give four-digit codes for the same character.

Three conditions must all be true for the keystroke to work. First, Num Lock must be on. Second, you must use the physical numeric keypad, not the number row above the letters. Third, the application must accept keyboard-generated character input, which rules out some full-screen games and sandboxed windows.

The mechanism runs through the keyboard layout, not the font. Windows sends a character code, and the font renders it. This is a useful distinction: if ¿ shows up as a box on screen, the input worked and the font is missing the glyph. Fonts that include Latin-1 supplement characters, which is nearly all modern system fonts, render it fine.

Quantified context: code points, ranges, and coverage

The inverted question mark sits at U+00BF, decimal 191, inside the Latin-1 Supplement block that spans U+0080 to U+00FF. The plain question mark is U+003F and the inverted exclamation mark is U+00A1. Spanish pairs the two inverted marks as sentence openers, so a well-formed Spanish question looks like this: ¿Cómo estás? The opening mark is a distinct character, not a rotated version of the closing one.

Windows-1252 encodes ¿ as byte 0xBF, and ISO-8859-1 uses the identical byte. That shared value is why the character survives round-trips between legacy Western encodings. UTF-8, the dominant web encoding, represents it as the two-byte sequence C2 BF. On a UTF-8 system, the decimal number 191 typed into a Unicode-aware tool still maps to U+00BF, which is why the same digits work in several environments.

Character support is effectively universal for this code point. The Latin-1 Supplement block is included in Unicode's first 256 code points, so any font with basic Western European coverage contains the glyph. Global web usage of UTF-8 exceeded 98 percent of surveyed pages by the mid-2020s, which means almost any page you paste the symbol into can store and display it correctly.

One professional caveat: alt codes and Unicode code points are different numbering systems that happen to agree for this character. Many characters diverge, so do not assume an alt code will produce the same symbol on every system. For ¿, the agreement is genuine: Windows-1252 position 191 and Unicode U+00BF both yield the inverted question mark.

Platform-by-platform methods compared

Different operating systems expose the character through different input models. The table below lists the main routes and their requirements.

Platform Method Requirements Notes
Windows, keypad Alt+0191 Num Lock on, numeric keypad Fastest route; leading zero required
Windows, no keypad Alt+Ctrl+? or Character Map Some layouts differ AltGr+? often produces ¿ on international layouts
macOS Option+Shift+? Any Apple keyboard Produces ¿ directly in most apps
Linux, X11 Compose, then ? and ? Compose key enabled Sequence depends on the compositor
iOS and Android Long-press ? Touch keyboard Picker offers inverted form
Word and Office Alt+0191 or Insert Symbol Numeric keypad or menu Autocorrect can also convert "??" in Spanish mode
HTML authoring ¿ or ¿ HTML context Named and numeric entities both valid

The macOS shortcut is worth memorizing because it needs no keypad: hold Option and Shift, then press the question mark key. The AltGr route on Windows international layouts is also reliable, but layouts differ by country, so treat it as a shortcut that works on many systems rather than all of them.

For HTML, the named entity ¿ and the numeric entity ¿ both render the character. Using the character directly in a UTF-8 document is simpler and now standard practice. The entities remain useful in environments that constrain encoding or in email templates that mangle raw multibyte characters.

Edge cases and failure modes

Several failure patterns account for most "alt code not working" complaints. Each has a specific cause and a specific fix.

Typing digits on the number row instead of the keypad: the row keys send scanning codes, not character values, so nothing appears. Num Lock switched off: the keypad produces navigation commands instead of digits. Omitting the leading zero: some systems interpret 191 in the wrong code page and insert a different symbol or nothing. Using a laptop with a hidden keypad: many 14-inch and smaller laptops emulate a numeric pad through Fn, which may or may not pass through the alt-code driver. Sandboxed or remote sessions: a virtual machine or remote desktop window sometimes intercepts Alt combinations as host shortcuts. Font substitution: the character is inserted but renders as an empty box, which is a display problem, not an input problem.

The hidden-keypad case is the most common on modern hardware. On laptops without a dedicated numeric cluster, pressing Fn plus the overlay keys sometimes works and sometimes does not, depending on the keyboard firmware. When it fails, Character Map or the Mac-style input method is more dependable.

A second counterintuitive edge case involves autocorrect. Word and Google Docs in Spanish language mode often convert a typed "?" preceded by an interrogative word into "¿" automatically, because Spanish typography expects the opening mark. This behavior can confuse users who think the alt code failed when the program actually substituted a different character on its own.

Encoding mismatches in older data pipelines are the third frequent culprit. Text saved as Windows-1252 and then read as UTF-8 turns ¿ into two garbled characters. If you see "¿" in a web page or database field, the cause is a mismatch between saved and interpreted encoding, not a broken keyboard or missing font.

Misconceptions and what the symbol is not

One persistent misconception treats the inverted question mark as a rotated glyph. It is not. The character is drawn separately, with its own design in each typeface. Typography software designs the opening mark to optically balance the closing mark rather than to overlap a rotated copy. Professional fonts render the two forms with matched stroke weight and terminal shapes.

A second misconception holds that the alt code works universally as "Alt+191." The four-digit form with the leading zero is the reliable Windows keypad sequence. Three-digit variants work in some environments and fail in others, which is why support forums give conflicting advice.

A third misconception treats the character as punctuation suitable for English. It is not standard English punctuation. Using it in English text reads as a typo or a stylistic flourish. The mark's function is specific to languages and orthographies that mark interrogatives at both ends, chiefly Spanish and, in some conventions, Galician and Asturian.

A subtler professional caveat concerns data entry. Systems that validate names or addresses against Western character sets sometimes reject ¿ or silently substitute a plain question mark. If you are building a form, test the character explicitly rather than assuming UTF-8 handling throughout the stack. The failure tends to appear only in legacy validation layers.

Practical actions that fix typing problems

Work through the following steps in order when the alt code does not produce the symbol.

Confirm Num Lock is on and test with a known alt code such as Alt+0169 for the copyright sign. If that fails too, the problem is the keypad or driver, not the specific code. Type the four-digit form, Alt+0191, holding Alt for the entire sequence. Release Alt only after the last digit. If you have no numeric keypad, use Windows Character Map: search for "inverted question mark," copy it, and paste. On macOS, press Option+Shift+? directly. On mobile, long-press the question mark key and select the inverted form from the popup. For documents that need the mark repeatedly, insert it once and use the application's autocorrect or text-replacement feature to map a trigger such as ";q" to ¿. For web content, confirm the page declares UTF-8 encoding and paste the character directly; fall back to ¿ if encoding cannot be controlled.

A useful threshold for writers: if you need the character more than about a dozen times in a document, set up a text shortcut rather than repeating the keystroke. The setup takes under a minute and removes the dependence on keypad hardware entirely.

Timeline of the character in computing standards

The inverted question mark's presence in standards is old enough that it predates the personal computer. The table traces the milestones that explain why the code is what it is today.

Period Standard or event Effect on the character
1963 ASCII published Plain ? at 0x3F; inverted form absent
1987 ISO-8859-1 standardized ¿ assigned to position 191
1991 Unicode 1.0 released ¿ placed at U+00BF in Latin-1 Supplement
1990s Windows-1252 adopted widely Decimal 191 became the familiar alt code
2000s UTF-8 overtakes legacy web encoding C2 BF byte pair replaces single-byte storage
2010s Mobile touch keyboards add long-press pickers Input becomes layout-independent

The timeline explains a common confusion. People who learned the character in the 1990s associate it with a single byte, while modern systems store two. Both views are correct within their encoding, and the alt code still works because Windows maps it through the code page rather than through raw bytes.

Limitations and what the code cannot do

The alt code is an input shortcut, not a rendering guarantee. It inserts the correct character value into the document, but display depends on the font and the encoding. A file that stores the character correctly can still show a placeholder box on a system lacking the glyph, though that is now rare on modern operating systems.

The method also depends on an active keyboard driver that honors Alt-numeric input. Web applications that capture keystrokes for shortcuts can suppress the sequence. Terminal emulators sometimes require a different configuration. In those environments, copy-and-paste or a Unicode input method such as Ctrl+Shift+U followed by BF on Linux is more reliable.

Finally, the alt code does not create meaning. Punctuation rules are language-specific, and the inverted mark belongs to particular orthographic traditions. Using the character correctly is a matter of following the conventions of the language you are writing, not of typing skill.

For anything touching language services, publishing, or software localization, treat character encoding as a systems problem rather than a keystroke problem. The keystroke fixes today's document; encoding discipline fixes the pipeline. That distinction matters more than any single shortcut.