When a person with intellectual disabilities has communication and fine motor impairments, multiple choice tests (MCTs) are useful if they are designed to allow some corrections and to measure the reliability of the answers. Here that evaluation method was applied to a home-based mathematical program, developed with a nonverbal adult with Autism Spectrum Disorder (ASD), Down syndrome (DS), and cerebellar damage. His counting abilities up until 21 were assessed with 185 MCTs, over 7 years (age: 22-28): he was able to count with a high accuracy (around 91%) from the beginning, independently from the numerosity, but influenced by his emotions. Even so, in 19 numerosity comparison trials he answered randomly and in 'giving N objects' (1-9) he failed several times. In the mental solution of 48 simple word problems with one operation, the overall accuracy was 59% for addition, 61% for subtraction, 67% for multiplication and 18% for division, without any previous training. These results confirm an unusual variability in the mathematical performances, already observed in both DS and ASD. Here his fine motor impairment hindered the possibility of writing and drawing without any help and only this use of MCTs revealed his knowledge. The student enjoyed doing mathematics.