BMAT停考后 UCAT四科时间与SJT分档
如果你还在搜 UKCAT，先记住一件事：这个考试在 2019 年就改名叫 UCAT，现在官网上的全部材料都用 UCAT 这个名字。另一个更容易踩的坑是 BMAT，它已经停考，最后一场在 2023 年 10 月，2024 年之后没有任何一场 BMAT。
很多学生以为申请牛津剑桥医学还得考 BMAT，于是去找 2023 年前的 BMAT 真题。花钱是小事，占掉的时间才是问题。这篇把两件事讲清楚：BMAT 为什么消失，UCAT 现在的四科怎么计时怎么打分，以及分数出来之后学校实际怎么用。
If you are still searching for the UKCAT, the first thing to fix is the name: the test was renamed the UCAT in 2019, and every current page on the official site uses that name. The second, more costly trap is the BMAT, which no longer exists. Its final sitting was in October 2023, and no paper has been set since. Plenty of students still buy BMAT past papers for Oxford and Cambridge, spend three weeks on them, and only then discover the test was retired. This article covers what replaced it, how the four current subtests are timed and scored, and what admissions tutors do with the number you receive.
BMAT 最后一场是 2023 年 10 月
BMAT 由 Cambridge Assessment Admissions Testing 承办，这家机构在 2023 年底停止提供入学考试，2024 年解散。2022 年官方已经宣布逐步停办 BMAT，理由之一是运营成本：只有七所医学院使用，却要维持独立的命题、印刷和考场体系。
2023 年 10 月那场是最后一次开考。2024 年起没有新卷，2025 和 2026 也没有补考安排。原本用 BMAT 的七所学校，也就是牛津、剑桥、帝国理工、UCL、利兹、布莱顿与萨塞克斯、兰卡斯特，从 2025 entry 起全部改用 UCAT，牛津在 2024 年 2 月确认了这一步。
常见的错误是以为改用 UCAT 只是换了个考试名字，备考内容可以接着用。BMAT 有物理、化学、生物的知识题和一小时论文写作，UCAT 一道学科题都没有，考的是语言推理、决策、数量推理和情境判断。两套题的训练方法几乎不重叠。
The BioMedical Admissions Test was run by Cambridge Assessment Admissions Testing, which stopped delivering admissions tests at the end of 2023 and was disbanded in 2024. The phase-out had been announced in 2022, partly for cost reasons: seven medical schools used the test, yet it needed its own item writers, printing and invigilation network. The last sitting was October 2023, and no replacement paper has appeared since.
All seven former BMAT schools now require the UCAT: Oxford, Cambridge, Imperial, UCL, Leeds, Brighton and Sussex, and Lancaster. Oxford confirmed the switch in February 2024, and the UCAT has been the single medicine test across the UK since the 2025 entry cycle. The content is the real trap, not the name. BMAT asked science questions and a written essay; the UCAT asks none of that, so a student who recycles BMAT chemistry revision is preparing for a paper that no longer exists.
VR 44 题 22 分钟 DM 35 题 37 分钟
2026 版的 UCAT 分四节，各自独立计时，每节前面还有 1 分 30 秒到 2 分钟不等的说明页。语言推理 44 题 22 分钟，平均每题 30 秒；决策 35 题 37 分钟，每题超过 1 分钟；数量推理 36 题 26 分钟；情境判断 69 题 26 分钟，平均每题 22 秒左右。全部做完约 117 分钟，接近两小时。
这个时间分配决定了训练重点。语言推理和情境判断是速度题，慢一秒就是少一题；决策和数量推理是计算与推理题，读完条件再动手才划算。很多学生以为四节可以平均用力，实际上去年不少人在语言推理上读得太细，最后十题全靠猜。
数量推理允许用屏幕上的计算器，这一点和 A-Level 数学的手算习惯不同。情境判断不计入总分，只给 Band，但它单独影响一部分学校的筛选，尤其是明确写明会看 SJT 的学院，报名前查一次目标校的表述。
The 2026 UCAT has four separately timed subtests, each preceded by a short instruction screen. Verbal Reasoning gives you 44 questions in 22 minutes, about 30 seconds each. Decision Making gives 35 questions in 37 minutes, so over a minute each. Quantitative Reasoning is 36 questions in 26 minutes, and Situational Judgement is 69 questions in 26 minutes, roughly 22 seconds per item. With instruction time, the standard test runs 117 minutes.
That split tells you where the pressure sits. Verbal Reasoning and Situational Judgement are speed tests: any second spent re-reading is a question you never attempt. Decision Making and Quantitative Reasoning reward reading the whole stem before calculating. Quantitative Reasoning permits an on-screen calculator, which surprises students trained to work mentally for A-Level Maths, and Situational Judgement is reported separately as a band rather than adding to the total.
三个认知科目各 300 到 900 分
语言推理、决策、数量推理三节各按 300 到 900 分换算，三节相加就是总分，区间 900 到 2700。原始分不能直接相加，因为三节题量不同，分别是 44 题、35 题、36 题。决策的多陈述题按 2 分计，部分正确给 1 分，这是唯一有部分分的题型。
2025 年起抽象推理被取消，总分才从 3600 降到 2700。查旧资料时看到 3000 分以上的成绩，那是 2024 年及以前的量表，不能直接和现在的分数比较。UCAT 官方在 2025 年的技术报告里说明，取消抽象推理是因为它的预测效度偏低，也更容易被短期训练拉高。
情境判断按 Band 1 到 4 报告，Band 1 最高。它不进总分，也不做加权平均。学生最容易把 SJT 当成性格测试来准备，去找所谓的标准人格答案，而实际打分的依据是专家小组对每个情境的判断，考的是你在具体场景里能不能抓住关键因素。
换算的细节要知道一点：量表分不是把正确率乘以六，而是按等值换算把不同年份、不同卷面的原始分放到同一条尺子上。UCAT 官方技术报告提到，为了让三科均分不至于越拉越开，近几个考季对语言推理上调、对数量推理下调过 10 到 20 分。同样答对八成，在语言推理和数量推理上换出来的分数并不相同。
Each cognitive subtest is scaled to 300-900, and the three scores add up to a total between 900 and 2700. Raw marks cannot simply be summed because the subtests differ in length: 44 items, 35 items and 36 items. Decision Making is the only subtest with partial credit, awarding one mark out of two for a partly correct response on multiple-statement questions.
Abstract Reasoning was removed in 2025, which is why the maximum fell from 3600 to 2700. Any guide quoting a score above 2700 is using the old scale and cannot be compared with a current result. Situational Judgement is reported as Band 1 to 4 with no contribution to the total; it is marked against an expert panel's judgement of each scenario, not against a personality profile, so rehearsing a correct character does not work.
Scaling is not a percentage conversion. Raw marks are equated onto a common 300-900 scale so results stay comparable across papers and years, and the technical report records deliberate adjustments in recent cycles, lifting Verbal Reasoning and trimming Quantitative Reasoning by 10 to 20 points to stop the three subtest means drifting apart. Two candidates with the same raw percentage in different subtests will not receive the same scaled score.
2026 考季 42,359 人 均分 1918
UCAT 官网公布的 2026 考季统计截止到 9 月 24 日，共有 42,359 人参加。认知总分均分 1918，三科均分分别是语言推理 602、决策 640、数量推理 675。第五个十分位，也就是中位数，是 1920；第九个十分位是 2250，换句话说总分 2250 以上才进入前 10%。
情境判断的分布更值得看一眼：Band 1 占 15%，Band 2 占 40%，Band 3 占 34%，Band 4 占 11%。拿到 Band 2 是大多数人的位置，Band 1 属于少数。一部分学校明确排除 Band 4 的申请，这条比总分更容易被忽略。
数字的意义在于定位，不在于制造焦虑。均分 1918 意味着刚过 2000 的分数已经在人群中间偏上，但离牛津的面试名单还有距离。牛津医学系公布的 2025 周期数据里，全部申请人均分 2217.5，进入面试的均分 2377.5，拿到 offer 的均分 2407.1。
The UCAT Consortium's 2026 statistics, valid to 24 September, cover 42,359 tests. The mean cognitive total was 1918, with subtest means of 602 for Verbal Reasoning, 640 for Decision Making and 675 for Quantitative Reasoning. The median sat at 1920, and the ninth decile, the top ten per cent, began at 2250.
Situational Judgement spread differently: 15 per cent of candidates were placed in Band 1, 40 per cent in Band 2, 34 per cent in Band 3 and 11 per cent in Band 4. Band 2 is the ordinary outcome, not a failure, though some schools exclude Band 4 applicants outright. Oxford's published 2025 cycle figures put the mean at 2217.5 across all applicants, 2377.5 for those interviewed and 2407.1 for those holding offers.
牛津把 GCSE 与 UCAT 等权 剑桥只看认知总分
牛津医学系的说法很具体：初筛把 UCAT 认知总分和 GCSE 的 8/9 比例放在一起做数值排名，两者等权。没有 GCSE 成绩、GCSE 少于五科，或者 2020 到 2021 年考 GCSE 的申请人，UCAT 权重翻倍。SJT 在这一阶段不使用。学校也写明没有固定分数线，因为每年面试人数固定，两个变量会互相补偿。
剑桥的口径更窄：只看认知三科的总分，2027 entry 明确不使用情境判断分数，也没有最低分门槛，分数用于决定面试名单，并可能在发 offer 时参考。这条对 SJT 拿到 Band 3 的学生是实际利好，但没有必要因此只申剑桥。
其余学校各有算法，有的设硬门槛，有的按比例加权，有的把 UCAT 和学术成绩做成综合分。口径每年调整，2026 年 9 月能查到的一版，到了 2027 年申请季可能已经换掉。目标校的 how we use the UCAT 页面，比任何二手排名表都可靠。
Oxford publishes its method in detail. Shortlisting builds a numerical ranking from the UCAT cognitive total and the proportion of GCSE passes at grades 8 and 9, with equal weight given to each. Applicants without GCSEs, with five or fewer, or who sat them in summer 2020 or 2021 have the UCAT double-weighted, and the situational judgement band is not used at this stage. Oxford also states there is no cut-off, since the number interviewed is fixed and the two variables compensate for each other.
Cambridge reads less: the overall cognitive total only, with the situational judgement score explicitly excluded for 2027 entry and no minimum threshold published. Other schools use thresholds, weightings or combined scores of their own, and these change between cycles. The current how we use the UCAT page on a school's own site is worth more than any third-party league table.
十四周节奏 前四周摸底后十周提速
把备考切成四段，每段三到四周。第一段做两套完整计时题，记录每节的正确率和未答数，这两个数字比总分有用。第二段按科目训练：语言推理练扫读和关键词定位，每天两组 44 题，重点是不回读；决策练逻辑结构和多陈述题的部分分策略；数量推理练估算和单位换算，把计算器用熟。
第三段回到整套计时，每周两到三套，做完必须复盘错题，写下错因，不是抄答案。第四段只做两件事：把最慢的题型压缩时间，把 SJT 的判断原则过一遍。学生最容易在最后两周继续刷新题，结果速度没提上去，错题也没消化。
一个具体的提速方法：语言推理的题干通常比选项长，先看问题问的是真、假还是无法判断，再回原文找那一句，不要在整段上平均用力。数量推理先看选项的数量级，能估出来的题不要按完整算式算。这些技巧不神秘，但需要至少二十套题才能变成习惯。
决策这一节要单独说：题型包括逻辑推理、论证分析、Venn 图和概率估算，多陈述题得先逐条判断五个陈述的真假，再决定选哪个组合，这样部分分至少能拿住。语言推理的题干分两类，一类问作者观点，一类问事实真假，两类的读法不同，混在一起练容易把节奏带乱。
A fourteen-week plan splits into four blocks of three to four weeks. The first block is diagnostic: two full timed papers, recording accuracy and unanswered counts per subtest rather than the total alone. The second block trains one subtest at a time, Verbal Reasoning through scanning and keyword location, Decision Making through argument structure and the two-mark multiple-statement format, Quantitative Reasoning through estimation and confident use of the on-screen calculator.
The third block returns to full papers twice or three times a week, with a written error log rather than a stack of finished answers. The final block cuts time on the slowest question type and reviews the principles behind Situational Judgement. Most students spend those last two weeks on fresh questions and finish no faster than they started.
Decision Making needs its own routine. Its items cover logical deduction, argument analysis, Venn diagrams and probability estimates, and the multiple-statement format rewards judging each statement separately before choosing an option, which preserves partial credit. Verbal Reasoning mixes opinion questions with factual ones, and the two need different pacing, so practising them as one block tends to blur both.
报名窗口 费用与四个常见失误
2026 考季的时间线是这样的：5 月 20 日开放账户注册，6 月 23 日开放考位预订，7 月 13 日开始考试，9 月 16 日 15:00 关闭预订，9 月 24 日是最后一天考试，10 月 15 日 18:00 是 UCAS 医学与牙医的截止时间，成绩在 11 月初发给学校。英国考场费用 70 英镑，英国以外 115 英镑，符合条件的英国考生可以申请 bursary 免掉。
四个常见失误：把账户注册当成预订考位，注册完就以为报上了；拖到九月才订，附近考场只剩边角时段；忘记带护照或驾照，考场不接受其他证件；同一年又报 UCAT ANZ，官方明确不允许两边都考。UCAT 一年只能考一次，成绩不能取消，也不能重考。
最后说一句方向性的问题：UCAT 只占申请的一部分，牛津的排名里 GCSE 等权，剑桥的面试和学术记录同样计入，个人陈述和推荐信也不是摆设。分数够用之后，把时间投到学术阅读和面试准备上，回报比再刷五套题更高。
The 2026 cycle ran to a fixed calendar: accounts opened on 20 May, booking on 23 June, testing from 13 July, booking closed at 15:00 UK time on 16 September, the last test day was 24 September, and the UCAS deadline for medicine and dentistry falls at 18:00 on 15 October, with results sent to universities in early November. The fee is 70 pounds in the UK and 115 pounds elsewhere, and eligible UK candidates can apply for a bursary that waives it.
Four operational mistakes recur: treating account registration as a booked test, leaving booking until September when nearby centres only have awkward slots, arriving without a passport or driving licence, and sitting both the UCAT and the UCAT ANZ in the same year, which the rules do not allow. The test can be taken once per cycle, and results cannot be cancelled or retaken. It is also only one input: Oxford weights GCSEs equally in its ranking, and Cambridge reads interviews and academic record alongside the score.
牛津高阶英汉双解词典 第10版 商务印书馆¥168📱 去京东购买京东精选
需要一对一辅导？Need One-on-One Tutoring?
A-Level / GCSE / IB / KS2-KS3 全科辅导
16621398022 同微信
Visit aleveler.com
或截图后用「扫一扫 → 相册」选择本图