This articles highlights some of the challenges associated with measuring human intelligence and in particular those associated with the measurement of the highest IQ. First, unlike measuring height, weight and even human speed on a 100 meter dash, measuring IQ (and particularly the highest IQ) is a difficult endeavor. Alfred Binet – one of the pioneering forefathers of modern IQ tests – pointed out that measuring human intelligence was not as clear cut as measuring other human traits and characteristics, and that it was therefore necessary to accept a degree of error in the measurement of IQ. This assertion was one of Binet’s finest. Another important concept in this debate is the diversity of IQ tests and associated scales. There are several different and well-respected IQ tests. Examples of well-known tests include Cattell Culture-fair IIIb, Weschler Adult Intelligence Scales (WAIS), Stanford-Binet, Woodcock Johnson, Raven’s Progressive Matrices to name a few. These individual tests have different constructs and are grounded in different although perhaps related theories on human cognition and might therefore measure different things (i.e. verbal or crystallized IQ vs. performance or fluid intelligence). Some tests are better at measuring certain types of human intelligence than others, while others do not test certain types of intelligence at all. For instance, Binet IV is one of the only well-known tests to have a quantitative reasoning scale. And subsequent revisions of the same test may change the measurement focus. Individuals may score more highly in some sections of the same test, while individuals may also score more highly on one test vs another. To make matters more complicated, several of the above-mentioned tests use the same mean score of 100 but employ a different standard deviations which makes the comparison of test scores meaningless unless these scores are adjusted for the standard deviation of the test in question resulting in the score being converted into a percentile. For instance, the standard deviation for the Woodcock-Johnson Test of Cognitive Abilities is 16. So a score of 132 (i.e. two standard deviations above the mean score of 100) would place a test taker in the top 2% of the population. The Cattell verbal tests on the other hand, have a standard deviation of 24, which means that a score of 148 (i.e. again two standard deviations above the mean of 100) corresponds to a score in the top 2% of test takers. Finally, IQ tests will normally have a ceiling (i.e. Cattell verbal has a ceiling for 161 for adults which corresponds to a result in the 99.48% of the population – or one in 192 people). But someone scoring 141 on the Woodcock-Johnson test would be in exactly the same percentile. This latter point makes estimating the highest IQ in the world very difficult as estimates are then required on top of the estimations and errors that are inherent in IQ testing.