All articles

The C-Test Format

One rule, applied 160 times: understand it properly and you have basically cracked the onSET before the first gap even appears. The onSET is a C-test, and a C-test runs on that one very friendly rule: a whole word is never deleted, only its second half. The beginning always stays. So you always see a stub, a stump of a word, and your job is to add the ending until a correct German word stands there again.

That is good news. You never start from zero. Every gap hands you a visible beginning and a complete sentence around it. From these two clues, stub plus context, you build the ending. In this article you will learn how a gap is made, what exactly stays visible, how points are awarded, and how to pace yourself.

Quick rule: Every gap is missing the second half of the word (the halving rule). Routine: read the stump → read the whole sentence → build the ending from context. 15 seconds per gap, never leave one blank.

How a C-Test Gap Is Built

The test author takes a short text and leaves the first sentence untouched, so you settle into the topic. After that the rule kicks in: starting from the second word, every second word loses its second half. The start of the word stays visible, the end disappears.

Watch how whole words turn into stubs. The visible part is the first half, the deleted part is the second:

Whole wordVisible (stub)Deleted (your job)
istist
TagTag
MenschenMenschen
trotzdemtrotzdem
KörperKörper
ZigarettenZigaretten

When the word length is even, the split is cleanly down the middle: Menschen has eight letters, four stay, four fall away. When the length is odd, you get one extra letter as a gift: Tag has three letters, two stay visible, only one is missing. So the visible part is never shorter than the deleted one.

What You See, What Is Deleted

On screen the stub sits fixed, and an input field blinks right next to it. You type only the missing ending, not the whole word. Think of it as a puzzle: the beginning is already placed, you attach the end.

Try it right away. The first sentence gives you the topic, then the gaps arrive. Read each stub, think the whole word, and type only the rest:

Viele Mens rauchen trot jeden Ta.

Did you feel it? You never guessed, you recognized. Mens in a context about smoking can hardly be anything but Menschen, so you add chen. trot calls up trotzdem at once, and Ta before jeden is a Tag. The stub narrows the options sharply, and the sentence does the rest.

The Halving Rule, Precisely

So you never doubt how much is missing, here is the rule in words: for a word with n letters, the first half stays visible, rounded up, and the second half is missing, rounded down.

Length nVisibleDeletedExample
321Ta / g
422da / nn (dann)
532dam / it (damit)
633Kör / per (Körper)
844Mens / chen (Menschen)

Remember just one thing: with an odd length, the visible beginning is the longer part. So you always see at least half of the word. Test your feel for the split on this sentence, where even and odd lengths are mixed:

Der Kör braucht da eine Pau.

Körper with six letters splits cleanly (Kör plus per), dann with four does too (da plus nn), and Pause with five gives you one extra letter to see (Pau plus se). The missing half is never longer than what already stands there.

Context Decides the Ending

Often the ending carries more than the rest of the stem: the grammar lives there. One stub can fit several endings, and only the sentence reveals which is meant. ein could become ein, eine, einen, or einem, yet the case in the sentence pins down exactly one form.

In this sentence, watch which ending the grammar demands, not just which word is meant:

Gestern ha ich ein interessanten Fi gesehen.

ha becomes habe because the subject is ich. ein becomes einen, not eine, because Film is masculine and stands in the accusative. And Fi in a context of sehen is clearly a Film. The ending here is no side detail, it carries the grammatical information.

Collocations and Fixed Phrases

The higher the level, the more fixed word partnerships help you. When wichtige stands next to Entscheidungen, the ending is almost given, because the adjective agrees with the noun. Such collocations, words that like to appear together, make longer words surprisingly easy.

This sentence sits at B2 level. Read the whole sentence first, then the endings almost fall into place:

Die Regie hat wich Entscheidungen getro.

Regie before hat calls up Regierung, wich before Entscheidungen demands wichtige (plural adjective), and getro at the sentence end is the participle getroffen, the fixed partner of Entscheidungen. You see: read the sentence as a whole and you barely have to search for the ending.

Scoring: One Point or No Point

Scoring is pleasantly simple and, importantly, free of anxiety. Each gap is worth exactly one point. There is no partial-credit system and, even nicer, no deduction for mistakes. A wrong or empty gap costs you only the one point you would not have earned anyway.

Your inputPoints
Correct ending1
Wrong ending0
Left blank0
Typo in the ending0

From this follows perhaps the most important strategic insight of this whole article: guessing costs nothing. Because there is no minus point, fill every gap, even when unsure. A reasoned guess always beats an empty field, because the empty field is a guaranteed zero, while the guess might earn the point.

Eight Texts, Rising Difficulty

One onSET run consists of eight short texts. Each text has exactly twenty gaps, so eight times twenty makes one hundred and sixty gaps in total. The texts do not carry a fixed level label, the task pool itself spans A2 to C1, but the difficulty noticeably rises from text to text: the early texts read easy, the later ones are denser, with rarer words and more complex sentences. The topic pool covers 18 areas, so cramming a single topic buys you little.

Scoring does not happen text by text, only the total score at the end counts. g.a.s.t. does not publish the exact cutoffs officially, but consistent sources put the approximate guide values at:

Total score (out of 160)Approximate level
up to 37below A2
38–60A2
61–91B1
92–124B2
from 125C1 (or higher)

The order is still your friend. Collect the safe points in the first, easier texts briskly, and grant yourself a little more calm for the later, denser ones. That way you never toss away easy points because time runs out at the end.

Your Time Budget: About 15 Seconds per Gap

Let us do quick math. Each text gives you a maximum of five minutes, so eight texts make forty minutes in total. Five minutes is three hundred seconds; divided by twenty gaps per text, that averages fifteen seconds per gap. That sounds tight but is not, because many gaps you solve in two or three seconds.

Once the five minutes run out, the next text loads automatically, and the last sixty seconds show a warning. If you finish early, confirm with "Text finished" and jump straight to the next text; the remaining time simply lapses.

The trick lies in the imbalance: invest the pace saved on easy gaps into the hard ones. If a gap stays shut after about ten seconds, type your best guess and move on. You can return at the end, but you must never bite down on a single gap while ten easy points wait unused.

A Quick Decision in Three Steps

When you get stuck on a gap, a small fixed routine helps. Three steps, always the same, so your mind does not tip into brooding:

  1. Read the stub. Which words even begin like that? Usually only a few remain.
  2. Read the sentence. What fits here in meaning and grammar? The context deletes the wrong candidates.
  3. Attach the ending. Write only the missing part and mind the grammatical ending.

Practice the routine on this C1 sentence. First a look at the stubs, then the whole sentence, then the endings:

Trotz de schlechten Wett blieben di Touristen optimi.

de before the genitive becomes des, Wett becomes Wetters, di before Touristen becomes die, and optimi at the sentence end becomes the long optimistisch. Even the long word at the close is tame once the sentence is clear: the stub optimi leaves hardly an alternative.

Worth remembering

The C-test deletes the second half of every second word; a little more than half of the letters stay visible (rounded up) and you supply the rest. You never start from zero, because stub and context lead you to the ending. Each correct gap earns a point, mistakes cost nothing, so leave no field blank. Eight texts with twenty gaps each make one hundred and sixty gaps in total, with rising difficulty from text to text, and about fifteen seconds per gap keeps you on schedule. Only the total score decides the approximate level.

One rule covers all 160 gaps, whether it is a verb ending, an article, or a rare noun. When you get stuck, the routine stays the same: read the stub, read the sentence, attach the ending. The next article shows you how to sharpen that routine under time pressure.

0/5