The 15 Hardest LSAT Logical Reasoning Questions of All Time

Are you ready to try the 15 hardest LSAT Logical Reasoning questions of all time?

Keep in mind that these questions present various types of difficulty. Some of the most difficult LSAT questions are so sly that you do not even realize you have done a difficult question—you confidently move on, only to later discover that you have missed the question. Other questions are clearly challenging from the start, and students knew they were stumped even as they were selecting an answer choice. So, this list presents some questions that were very clearly challenging on test day and others whose difficulty might have become more obvious only in retrospect.

The questions have been selected based on statistics from students’ practice tests, the number of student questions we receive, and our own experience teaching and writing about these questions. At the end, we’ll talk about some of the interesting patterns we noted within the list of 15 questions.

October 1994, Section 1, #24: Vinland Map Titanium Ink

Flaw in the Reasoning. This question involves the presence of titanium in both the Gutenberg Bible and the B-36 Bible. Respondents answered this question correctly at a rate of just 15%, meaning that the correct answer actually was harder to find that if one just guessed randomly (which would yield a 20% hit rate). From a statistical standpoint, this is the second hardest LSAT Logical Reasoning question ever to appear, although from a pure difficulty standpoint this may be the hardest.
December 1994, Section 1, #18: Grand Banks Cod

Must Be True. Must questions are not generally among the hardest Logical Reasoning questions, but the Grand Banks Cod question is one of the exceptions. The question contains a discussion of separate estimates of cod stocks, one of which has increased by the same amount that the other has decreased. The two estimates offset each other, leading the overall official estimate to stay the same. The very casual language in the correct answer makes this question even more challenging.
December 1995, Section 2, #23: Airline Seat Cancellation

Justify-PR. This was one of the first Justify-PR questions. In this argument, Arnold and Jamie argue about the moral obligation of an airline to compensate a traveler for a cancelled flight. Because of the Principle aspect in this question, the answers have a more abstract quality, and students answered it correctly at a rate of just 21%.
October 1996, Section 1, #16: Brown Dwarfs

Assumption-FL. There have been other questions about brown dwarfs on the LSAT before, but this was the first and by far the most difficult. The Formal Logic aspect by itself makes the question much harder for most students (most people probably think there’s a special level of hell reserved for the makers of Formal Logic questions). The addition of several science terms (brown dwarfs, red stars, hydrogen, lithium, helium) adds to its difficulty.
June 1997, Section 1, #19: Meteorite Impact Crater

AssumptionX. This is a tricky question, because the correct answer, which is not an Assumption of the argument, is incorrect because of the presence of one word: “more.” And, because the correct answer is (A), many students quickly bypassed (A) and then never returned to it, leaving them to choose an incorrect answer. The science element also adds to the difficulty of this problem.
June 1997, Section 2, #24: Provincial Taxes Weaken.

The stimulus in this problem discusses a proposal to stimulate economic activity by refunding $600 million in taxes, a proposal that the author calls “an illusory benefit” (that pretentious description was probably a sign that the problem itself wasn’t going to be easy). The question stem is also a bit unusual, increasing the difficulty of this problem.
October 1997, Section 2, #10: Nonmoving Vehicles

Strengthen. This question is one of the most frequently asked about questions on our LSAT Hotline. The stimulus is short and seemingly easy to understand. But the correct answer—which gives this problem its name—is difficult to understand, and most students dismiss it without much consideration when working through the choices.
December 1998, Section 1, #23: Coffee and Insomniacs

Flaw. This is an unusual problem in that many LSAT experts don’t think that this question is among the toughest questions ever. But, of course, this list isn’t for us, and students seem to routinely slip past the correct answer and fall for one of the several attractive incorrect answers provided
June 2001, Section 2, #25: Business Investment and Environmental Responsibility

Parallel Reasoning. In terms of size, this is the longest question on the list. And as this was a Parallel question positioned at the very end of the section, many students simply gave up when they reached the end and saw this monster looming there. Those factors led to a correct response rate of a mere 14%, the lowest of any question on this list. Overall, the question probably isn’t as hard as the numbers appear to indicate, but it is still a tough question, and the abstract nature of the Parallelism required warrants the inclusion of this question among the all-time hardest.
October 2001, Section 1, #23: Total Set of One’s Beliefs

Flaw. Tests throughout the 2000s have routinely presented questions with abstract moral discussions in the stimulus. Many of these questions have been extremely difficult. Topics such as good will, trust, and happiness have figured prominently in these questions, which have all probably been produced by the same test writer. We really wish he would stop writing questions! This question, an early example of his style, discusses accepting or rejecting beliefs on the basis of evidence, and students correctly answered this question at a rate of just 19%.
October 2002, Section 2, #16: Confidence, Trust, and Abilities

Justify. In our discussion of question #10 on this list (in Part II), we mentioned an LSAT writer who tends to put abstract moral discussions into the stimulus. This question appears to have been produced by that same writer, and addresses the issues of confidence, trust, distrust, abilities, and difficult tasks. Almost invariably, the questions from this author are tough, and this one is clearly no exception.
June 2003, Section 2, #17: Alex’s Capabilities

Parallel Reasoning. This question is probably another one from our friend who wrote questions #10 and #11 on this list, although this question’s stimulus is a bit easier to understand than the stimuli in those other two problems. However, it is paired with a Parallel question, which in this case increases the level of difficulty. Ultimately, only 17% of test takers were able to answer this question correctly.
June 2004, Section 1, #15: Cigarette Advertisements

Resolve. In this question, the author argues that governments have the right to ban cigarette advertisements, but that they shouldn’t because other unhealthy practices are already legal to advertise. Resolve questions are generally fairly easy, but students correctly answered this one at a rate of only 20%.
December 2004, Section 1, #12: Fish, Paper Mills, and Dioxin

Weaken. This infamous question wreaked havoc on test takers back in 2004. Only 21% of students answer this question correctly, and even LSAC’s discussion of the correct answer required two pages of explanation. It’s not in our character to hate LSAT questions, but this one would be near the very top of the list if we did.
June 2005, Section 2, #24: The Existence of Money

Justify. This may be the easiest question on this list, which goes to show how hard the other questions on this list are. The author in this stimulus asserts that money doesn’t exist, namely because all that is needed to make money disappear is for all of us to stop believing in it. Uhh, ok.

One final note about this list: by definition, any such list such is somewhat subjective, and a question that is a challenge for one student may not be difficult for another. The questions listed above are ones that students regularly ask about, and the ones we know from statistics and our experience to be extremely difficult. If you can correctly solve any of the questions on the list in less than two minutes, you have done a great job.

Observations

Now that we’ve completed our list of the 15 hardest LSAT Logical Reasoning questions, let’s get to the most valuable and interesting part of this discussion: what we can learn from these questions. Here are a few fascinating observations drawn from an analysis of the 15 Hardest LSAT Logical Reasoning questions:

1. The spread of the answer choices is notably biased towards (A)

Questions within this set of 15 are overwhelmingly biased in favor of answer choice (A). Certainly this could perhaps be partially explained by the fact that a list such as this one is somewhat subjective, but consider that the correct answer for the next five questions we considered for the list but did not include was also (A).

Overall, the correct answer for 15 questions breaks down as follows:

(A) = 9, or 60%
(B) = 1, or 6.6%
(C) = 4, or 26.6%
(D) = 1, or 6.6%
(E) = 0, or 0%

Again, keep in mind that if we had added the next five most difficult questions, creating a list of 20 rather than 15, answer choice (A) would have been correct 14 out of 20 times, or 70%! Given that random chance would suggest that each correct answer appears at an average rate of just 20% and that overall on LSAT Logical Reasoning answer choice (A) is correct just 18.7% of the time, what is going on?

First, it’s not an accident. The statistics are too far out of kilter to be random, and they clearly show that this is the deliberate result of an intentional strategy used by the test makers. Second, it shouldn’t be that surprising, and in the LSAT Logical Reasoning Bible we discussed this exact phenomenon. Often in the case of a particularly hard stimulus or question stem, students can become so confused that when they hit answer choice (A) they really aren’t in tune with the problem yet, and they tend to move on without being certain of what (A) was really about. At that point, they often get caught up in the later answer choices, and usually select one of those.

So, the takeaway here is that if you are really stumped (meaning you know that a particular question is extremely difficult) and you have to guess on one of these tough problems, guess answer choice (A). And keep in mind that if you are reading a tough stimulus, you should not discount (A) too quickly—it may very well be the correct answer!

2. The spread of the answer choices is notably biased away from (E)

Another interesting point bears discussion as well: why was (E) not the correct answer for any of the questions on this list? Probably because of what we call the “reverberation effect.” Often, the very last thing a student reads “reverberates” in the memory, making one more prone to choose that last answer if no other answer has immediate appeal. With the questions on this list, which are specifically designed to be challenging, the test makers probably don’t want the reverberation factor leading students to the right answer, so they avoid making (E) the correct answer choice.

3. The questions on this list tend to show up late in the section.

The placement of these questions does not appear to be random. The range of question placement for the Top 15 Hardest questions is as follows:

Average Question Position: 20.33

Of the 15 questions, 0 appear in the single digits, 5 of them appear in positions #10-#19, and the remaining 10 appear from #23-#25. Note that last number: 2/3 of the Top 15 appear in the range of #23-25, the exact point that many great test takers are nearing the end of both time and energy. Thus, if you are shooting for a top score, at the very end of the section you must be careful of a subtle, sinister attack on your raw score: many of the test makers’ most brutal attacks will come at a point when most test takers are already worn down. And, if you are just reaching the end and hoping to optimize your score, be wary of anything that even feels tough from the outset.

4. Watch the question types

The question types of the Top 15 also tend to be non-random:

Flaw
Must
Justify-PR
Assumption-FL
AssumptionX
Weaken
Strengthen
Flaw
Parallel
Flaw
Justify
Parallel
Resolve
Weaken
Justify

Must = 1
Flaw = 3
Parallel = 2
First Family = 6, or 40%

Strengthen = 1
Assumption = 2
Justify = 3
Resolve = 1
Second Family = 7, or 46.66%

Weaken = 2
Third Family = 2, or 13.33%

While this may appear to be a fairly random selection of question types, if you remove the pure fact based questions—such as Must and Resolve—the question types that involve more abstract reasoning account for 13 of the 15 questions, or 86.66% (These more abstract question types include Flaw, Parallel, Strengthen, Assumption, Justify, and Weaken questions). Thus, as would be expected, the problems that rank among the hardest of all time tend to be most commonly based on more abstract questions.

5. Science and philosophy rule

From a topical standpoint, 5 of the 15 questions were based on science topics, and 4 of the 15 were based on more abstract philosophical-style stimuli (as exemplified by question #10 on this list). This isn’t to say that every question on one of those two topics is sure to be difficult, but as you get deeper in the section, try to be a bit more deliberate when facing questions based on those two topics.

This ends our discussion of some of the most difficult Logical Reasoning questions of all time. We hope you find the list useful, but more importantly, we hope the analysis at the end provides you with some helpful insights into the LSAT. Again, keep in mind that this list is subjective, and you are welcome to add your personal least favorite LSAT questions below—we considered a many challenging questions for this discussion, and there are obviously a lot of extremely difficult questions that didn’t show up on this list.