The Ultra Test


Change History:

02/06/24 Scoring fee decreased to $10.
09/24/01 Scoring fee increased from $20 to $30.
04/05/01 Scoring fee decreased to $20.
01/04/00 Scoring fee increased from $25 to $33.
6/8/97 Reverted back to the drawings of dodecahedra and icosahedra that Hoeflin uses, which may be clearer.



The Ultra Test is an untimed, unsupervised intelligence test for gifted adults that is intended as an admissions test for various high-IQ societies, of which the test designer founded four -- Mega, Prometheus, the One-in-a-Thousand Society, and the Top One Percent Society -- and currently serves as editor and sole officer for the latter two. These groups have percentile requirements at the 99.9999, 99.997, 99.9, and 99th percentiles, respectively. The test designer's Mega Test was published in the April 1985 issue of Omni magazine and his Titan Test in the April 1990 issue of that publication. Both of those tests are still in use, but the test designer hopes eventually to combine the best problems from all three tests to form one or two higher-quality tests. (In fact, one of these tests, the Hoeflin Power Test, has already been compiled. -- DTM)



1. ANSWER SHEET. Please fill out the answer sheet as instructed. Provide the other information requested too.

2. TIME LIMIT. There is no enforceable time limit, but 100 hours spread over one or two months proved to be the optimum for the Mega and Titan tests.

3. ADDITIONS AND CORRECTIONS TO ANSWERS: No additions or corrections to your initial set of answers will be accepted. You get only one try at this test, so do your best the first time.

4. ASSISTANCE. So that no one will gain an unfair advantage by using reference aids, everyone is both permitted and encouraged to use books (In response to the question of online library catalog searches, Hoeflin writes: "I'd prefer that people employ just a dictionary and thesaurus in book form ... my test is not intended to be a test of computer-using skills."). Paper and pencil should suffice for this test, but if desired, you may use a non-programmable pocket calculator (for addition, subtraction, multiplication, or division); however, computers are not allowed as aids to the test. This restriction includes, but is not limited to the use of: commercial offerings such as spreadsheets and mathematical programs, user-written programs, electronic encyclopedias, Web search engines, and online sequence databases. Any assistance from other persons is prohibited.

If you have plans to take Hoeflin's Power Test (a compilation of the best nonverbal and non-sequence problems from the Mega, Titan, and Ultra Tests), you should be aware that that test has more restrictive rules, disallowing the use of calculators and reference material altogether. I have provided cross-references to the Power Test -- you are free to use or ignore these links as appropriate for you. [DTM]

5. DISCUSSION OF PROBLEMS. As the Ultra Test may be used as an admissions test to several societies, please do not share answers in a public forum with anyone who has not tried this test already. If you have plans to take Hoeflin's Power Test, then you should not share answers at all.

6. GUESSING. There is no penalty for wrong answers or guesses, so it is to your advantage to guess whenever you are unsure of an answer. Verbal items count one (1) point and nonverbal items count two (2) points, so your maximum raw score would be 72 and your minimum score would be zero (0). Misspellings on the verbal problems will be given half credit if otherwise correct.

7. FEE. Email answer sheets to The scoring fee for each test is $10 USD, payable via Stripe. You will receive a scoresheet showing your raw score, your corresponding I.Q. score, and its estimated percentile in the general population.

8. SCORESHEETS. Send to the address above. Allow up to four weeks to elapse before complaining about not receiving a score report. Most scores are sent out within two or three weeks of the receipt of your answers.


A comment on verbal intelligence: Some people wonder about the value of verbal items as a measure of "intelligence," since trying such items seems to involve little or no intellectual effort. From the purely statistical standpoint, many studies repeatedly showed that verbal intelligence, including the sheer size of one's vocabulary, has one of the highest correlations of any type of test item with overall intelligence as measured by tests containing a wide variety of test items. See for example the book Intelligence in the United States, published around 1958, for ample documentation. On the purely intuitive level, one might say that learning a language, including vocabulary, is for the child like decoding hieroglyphics. The brighter child will master this decoding process far more readily than the average child. Later, of course, one can artificially boost the size of one's vocabulary, but cleverly designed tests of verbal intelligence can get around this problem by relying on somewhat atypical verbal items that one would be unlikely to pick up through a "vocabulary improvement" course but that a gifted child would be likely to have picked up if he has been reasonably inquisitive -- and isn't inquisitiveness an important part of intelligence? Finally, to use a computer analogy, a powerful computer without adequate software (analogous to verbal intelligence in humans) would be relatively unproductive no matter how powerful the hardware. -- RKH



Find the best solution for each of the following analogies. Example -- CLEAR : CLOUDY :: TRANSPARENT : ? The best answer would be TRANSLUCENT. Misspellings will be given half credit if otherwise correct. The final authority, wherever possible, for correct spellings will be the Merriam-Webster Collegiate Dictionary, 10th edition.

1. Once : Twice :: Bitten : ?
2. Amphibian : Salamander :: Political district : ?
3. Riddle : Mystery :: Mystery : ?
4. Penny : Thrift :: Pinch : ?
5. 2.54 : Inch :: 454 : ?
6. One-eyed : Cyclops :: Two-faced : ?
7. Swiftness : Velocity :: Stickiness : ?
8. Say : Hear :: Imply : ?
9. Space : Hyperspace :: Vector : ?
10. Wind : Rain :: Typhoon : ?
11. Scenic : Picturesque :: Roguish : ?
12. Inward : Outward :: Infection : ?
13. Strong : Herculean :: Polymorphous : ?
14. Sophisticated : Wisened :: Wrinkled : ?
15. Wicked Woman : Witch :: Bad Taste : ?
16. Silly, Not Obese : Fatuous :: Offensive, Not Loud : ?
17. Column : Row :: File : ?
18. Humbug : Bach :: Seek : ?
19. Coals : Newcastle :: Rough Beast : ?
20. Enlightenment : Illuminati :: Knowledge : ?
21. Pride : Prejudice :: Sense : ?
22. Of Ten : Factor :: Of Magnitude : ?
23. 2.54 : Inch :: 3.26 : ?
24. Pocus : Hocus :: Pokery : ?
25. Eggs : Grading :: Wounded : ?
26. Mock : Mach :: Oiler : ?
27. Go : Gang :: Awry : ?
28. Tall, Dark : Handsome :: Nasty, Brutish : ?
29. Split Apart : Cleave :: Stick Together : ?
30. Image : Idea :: Hallucination : ?
31. Hairpiece : Wig :: Party : ?
32. Tom : Harry :: Gold : ?
33. Them : Us :: Eskimo : ?
34. Wedding Assistant : Best Man :: Movie Production Assistant : ?
35. A, AB, B, BO, O : BO :: A, C, G, T, U : ?
36. Plus Ultra : Ne :: Ne Sais Quoi : ?



37. If the four sides of a square consist of rods each of which is painted white or black, six distinct color patterns are possible: (1) all sides white, (2) all sides black, (3) one side white and the rest black, (4) one side black and the rest white, (5) two adjacent sides white and the other sides black, and (6) two opposite sides white and the other sides black. Suppose that each of the twelve edges of a cube is a rod that is painted white or black. How many distinct patterns are possible if any three of the rods are painted white and the other nine are painted black?
38. Suppose that the figure at right consists of nine rods of equal length joined together to form four equilateral triangles of equal size. If two of the rods are painted white and the remaining seven rods are painted black, how many distinct patterns can thereby be created?
39. Suppose an octahedron consists of twelve rods all of equal length and forming eight equilateral triangles -- the eight sides of the octahedron. If any two of the rods are painted white and the rest black, how many distinct patterns are possible?
40. Suppose that the figure at right consists of thirty rods of equal length that form twelve pentagonal figures of equal size, which form the twelve sides of a regular dodecahedron. If any two rods are painted white and the remaining twenty-eight are painted black, how many distinct patterns are possible? [cf. Hoeflin Power Test, problem 30]
41. Suppose an octahedron consists of twelve rods all of equal length and forming eight equilateral triangles -- the eight sides of the octahedron. If any three of the rods are painted white and the rest black, how many distinct patterns are possible?
42. If lightbulbs are put at two different corners of a square, two distinct patterns are possible: one in which the bulbs are at opposite ends of any side of the square, and one in which the bulbs are diagonally across from one another. If lightbulbs are put at four different corners of a cube, how many distinct patterns are possible?
43. If lightbulbs are placed at two different vertices of a regular dodecahedron, how many distinct patterns are possible? [cf. Hoeflin Power Test, problem 5]
44. Suppose that lightbulbs are placed at any three distinct vertices of a regular icosahedron (illustrated at right). How many distinct patterns can thereby be formed? [cf. Hoeflin Power Test, problem 6]
45. The five figures shown below represent the appearance of a solid, opaque object as seen from five of its six sides. Each line shown depicts a side of the object that is perpendicular to the plane of this page. The object was constructed by gluing together a number of identical cubes so that at least one face of each added cube precisely and entirely covers and is everywhere contiguous with one face of a previous cube. Draw the sixth view of the object.


The following number sequences are each based on pi, whose first fifty-one digits (counting the initial 3) are as follows: 3.14159 26535 89793 23846 26433 83279 50288 41971 69399 37510. Find the number that best continues each of these sequences:

46. 3 4 5 5 7 5 1 9 1 8 9 ?
47. 2 2 3 2 4 10 1 7 4 4 4 ?
48. 8 6 10 15 16 17 13 14 13 ?
49. 6 13 15 31 39 43 45 ?

Draw the figure that should fill the blank (identified by the question mark) in each of the following series: [cf. Hoeflin Power Test, problems 32 through 35]

This concludes the test.


Look at the Ultra Test norm.

Republished from Darryl Miyaguchi's hosted version (unmaintained since 2004).