Do I need all 12 patterns to score on Part 5?

**No — locking the top six at 90%+ already covers ~65%**. The bottom six items appear at 4-6% each, so beginners should **concentrate on the top six**. **Trying to perfect all 12 simultaneously usually drops top-six precision** instead of raising the floor.

What is the POS-identification trap?

The classic trap is **assuming "article + adj + noun" when the structure is actually noun-noun compound**. "the customer ___ team" — **service (noun)** is correct, forming **customer service team**. **Heuristic**: when both **service (noun)** and **serviceable (adj)** are options, judge by sentence sense — "service team" vs "serviceable team".

When to use that vs which?

**that = restrictive only; which = both**. **The hard rule: after a comma (non-restrictive), "that" is forbidden**. **"the company, ___ we invested in" → which (correct), that (wrong)**. Without a comma (restrictive), both work. **TOEIC Link tests non-restrictive "which" frequently**.

How to handle tense items with no time marker?

**Read meaning + verb nuance**. **"The company has been growing" (continuous) vs "The company grew" (closed)** — pick from continuity / completion / progressive cues. **About 30% of Part 5 has no overt time marker**, and statistical bias favors **simple present / simple past (60-70%)** over perfect / progressive in those slots.

Is 5-6 weeks to 80% on Part 5 grammar realistic?

**Yes, if you sustain ~30 minutes / day**. The weekly cycle is **(1) one of the top six patterns per week — "core image + 30-item drill". After 6 weeks, all six covered. (2) 50 past-paper items as field test. (3) wrong-answer log + reinforcement.** **Reaching the "pattern recognition" stage is the precondition** — pure memorization tops out near 70%. **EnglishBlitz Grammar drills are designed to accelerate exactly this pattern-recognition step**.

TOEIC LinkPublished May 2, 2026

TOEIC Link Grammar — 12 Most-Frequent Error Patterns and the Study Order to Reach 80% on Part 5

TOEIC Link Part 5 (incomplete sentences) grammar items collapse into about 12 error patterns, and the top 6 patterns cover ~70% of all grammar items. Studying every grammar topic exhaustively is far less efficient than recognizing the high-frequency patterns as templates. This guide profiles all 12, the typical wrong-answer structures, the matching grammar references, and a study order that gets Part 5 grammar to 80%.

Structural map of Part 5 grammar items

Part 5 = 30 items with vocabulary and grammar items mixed roughly 60:40. Of the ~12 grammar items, the patterns below cover the bulk.

Top 12 patterns (Volume 1-3 + 100 mock sets): 1. POS identification (verb / noun / adj / adv) — 18% 2. Verb tense (present / past / perfect / future) — 15% 3. Subject-verb agreement — 9% 4. Relative pronouns (which / who / that / whose) — 8% 5. Preposition choice (in / on / at / for / by …) — 8% 6. Conjunctions (although / despite / however / because …) — 7% 7. Passive voice (active / passive / perfect-passive) — 6% 8. Comparatives / superlatives (more / -er / most / -est) — 5% 9. Subjunctive (if / had / were / would) — 5% 10. Infinitive vs gerund (to + V / -ing) — 5% 11. Participle phrases (-ing / -ed modifiers) — 4% 12. Inversion / emphasis (only / never / not only) — 4%

Top six (POS / tense / S-V / relatives / prepositions / conjunctions) cover 65%. Locking those at 90%+ secures 8/12 = 67% of Part 5 grammar.

The remaining four items appear at 4-6% each. Study time should follow frequency, not surface uniformly.

Part 5 grammar ≈ 12 items
Twelve patterns to recognize
Top 6 patterns = 65% of items
POS most frequent (18%)
Tense + S-V together: 24%
Lower 6 patterns: dispersed

Patterns 1-3 — POS / tense / agreement (32%)

Pattern 1 — POS identification. Pick verb / noun / adj / adv from slot context. Typical: "The committee made a ___ decision" → adjective (final / important / quick). article + ___ + noun ⇒ ~90% adjective. Trap: "article + noun + ___ + noun" is a noun-noun compound (e.g., customer service), so the answer is a noun, not adjective.

Pattern 2 — Verb tense. Time markers (yesterday / next week / since 2020 / by tomorrow) are the cue. since 2020 → present perfect, by tomorrow → future perfect, yesterday → past simple maps mechanically. Exception: conditional clauses use present even for future (If it rains tomorrow, …).

Pattern 3 — Subject-verb agreement. Identify singular vs plural subject. Typical: "The team of engineers ___ working on the project." → "is" (singular). Pitfalls: collective nouns / Latin plurals — team / committee / staff / data / criteria. The contrast the number of ___ is / a number of ___ are also recurs.

Recommended study order: POS → tense → S-V. POS is the most mechanical and the steady scoring source, while tense and agreement need full sentence-meaning processing.

POS: article + ___ + noun
Noun-noun compound trap
Time marker → tense map
Conditional clause exception
team / data / criteria pitfalls
POS → tense → S-V order

Patterns 4-6 — relatives / prepositions / conjunctions (23%)

Pattern 4 — Relative pronouns. Pick by antecedent (person / thing) + case (subject / object / possessive). Person-subject = who, thing-subject = which, either-subject = that, possessive = whose. Rule: "comma + ___" (non-restrictive) disallows that. Typical wrong: "the company, that we invested in" → correct: "the company, which we invested in".

Pattern 5 — Preposition. Pick by time / place / means / purpose context. Time: in (year / month) / on (day) / at (clock) | Place: in (large area) / on (surface) / at (point) | Means: by (transport, comms) / with (tool) / in (language) maps mechanically. by Friday (deadline) vs until Friday (continuation) — high-frequency contrast.

Pattern 6 — Conjunctions. Pick by logical relation (concession / addition / cause / purpose). although vs despite (clause vs noun phrase), because vs because of (clause vs noun phrase), so that (purpose) vs so + adj + that (result). conjunction vs preposition vs conjunctive adverb distinctions are heavily tested. However is a conjunctive adverb — needs a semicolon mid-sentence.

High-yield wrong answer: "despite + clause", "because of + clause" are always grammatically wrong. 70% of learners miss this — high priority.

Relative: antecedent + case
No "that" after a comma
Time / place preps mechanize
by deadline / until continuation
although clause / despite phrase
However = conjunctive adverb

Patterns 7-12 — passive / comparison / subjunctive / others (35%)

Pattern 7 — Passive. Determine if subject is the receiver of the action. "The proposal was approved by the committee" is canonical. Perfect passive (have been + p.p.) and progressive passive (be being + p.p.) appear at the higher score band.

Pattern 8 — Comparatives / superlatives. than → comparative, the → superlative. Memorize regular (-er / -est) and irregular (good-better-best, bad-worse-worst, many-more-most). Idioms: twice as ___ as / -er than ever / the most ___ in / of.

Pattern 9 — Subjunctive. Pick by if-clause tense. Past (present hypothesis) / past perfect (past hypothesis). If I were you / If she had known / I wish I had as reflex idioms. Part 5 leans toward present hypothesis (If + past / would + base).

Pattern 10 — Infinitive / gerund. Verb-specific selection. enjoy / consider / suggest → gerund | want / hope / decide → infinitive | remember / try / stop → both (sense changes). A 30-verb table covers ~90% of the cases.

Patterns 11-12 — Participles / inversion. -ing / -ed sentence-modifying / Only after / Not until inversion — the 350+ separator items. Below 350, you can deprioritize them.

Passive: subject = receiver
Comparison: than / the cue
Subjunctive: if-clause tense
Gerund vs infinitive: 30-verb table
Participle / inversion: 350+ separator
<350 target: deprioritize

Reference grammars and study order

Beginner (200-250 target). Ichiokunin no Eibunpou (Onishi Yasuto) for core image, plus TOEIC L&R Bunpou Mondai Deru 1000 Mon (Narishige Hisashi) for high-frequency drills. Core image + drill is the most retentive 2-stage combo for beginners.

Intermediate (250-300 target). Grammar in Use Intermediate (Cambridge) for native explanatory framing, plus EnglishBlitz Grammar drills tuned to TOEIC Link. The English-language reference captures nuance that Japanese explanations sometimes miss.

Advanced (300+ target). Practical English Usage (Michael Swan) as a dictionary, plus decomposing every missed grammar item from past papers. Volume of drilling > linear reading at this stage.

Recommended sequence. (1) Top six patterns (POS / tense / S-V / relatives / prepositions / conjunctions) — 1 lap in 2-3 weeks. (2) 200 past-paper items in field test. (3) Track wrong-answer patterns and reinforce the bottom six (passive / comparison / subjunctive / inf-gerund / participle / inversion) over 2-3 weeks. (4) 300 past-paper items to set. 5-6 weeks lands Part 5 grammar at ~80%.

Anti-pattern: linear cover-to-cover grammar reading. You waste cycles on TOEIC-Link-irrelevant constructions (advanced inversion / cleft / dummy subject) and the high-frequency precision never sharpens.

Beginner: Onishi + Deru 1000
Intermediate: Grammar in Use + EB
Advanced: Practical English Usage
Top 6 patterns first
5-6 weeks → 80%
Linear reading is inefficient

Top-12 grammar pattern frequency map

Rank	Pattern	Frequency	Score band	Priority
1	POS identification	18%	200-300	Top
2	Verb tense	15%	200-300	Top
3	Subject-verb agreement	9%	200-250	High
4	Relative pronouns	8%	250-300	High
5	Preposition	8%	200-300	High
6	Conjunctions	7%	250-300	High
7	Passive	6%	300+	Mid
8	Comparison	5%	250-300	Mid
9	Subjunctive	5%	300+	Mid
10	Infinitive / gerund	5%	250+	Mid
11	Participle	4%	350+	Low
12	Inversion / emphasis	4%	350+	Low

* Frequency aggregated from Volume 1-3 + 100 mock sets. Score band is the median match from EnglishBlitz internal learning data.

Six rules for grammar pattern attack

Treat the 12 patterns as templates
Top six = 65% coverage
POS is the most frequent (18%)
Allocate time by frequency
5-6 weeks → 80% on Part 5 grammar
Avoid linear cover-to-cover reading

Drill grammar in EnglishBlitz Read the Reading strategy

Frequently Asked Questions

TOEIC® and TOEIC Link™ are registered trademarks of ETS. EnglishBlitz is not affiliated with, endorsed by, or associated with ETS. Frequencies, score bands, and time budgets reflect EnglishBlitz internal analysis and official-book aggregation; specific test forms and learners may vary.