The FRAXA trinucleotide repeat at Xq27.3 gives rise to fragile X syndrome when fully expanded, and both premature ovarian failure (POF) and fragile X tremor and ataxia syndrome (FXTAS) when in the premutation range. Reports of phenotypic effects extending into the intermediate repeat range are inconsistent but some studies suggest that these smaller expansions predispose to special educational needs (SEN). This study utilises the Avon Longitudinal Study of Parents and Children (ALSPAC) cohort to investigate cognitive and behavioural variables that might be associated with FRAXA intermediate alleles. The current study failed to find any strong evidence of association of FRAXA intermediate alleles with SEN, behavioural problems or cognitive difficulties. However, our findings illustrate some of the difficulties encountered in identifying individuals with SEN. The power to identify specific components of cognitive and behavioural difficulties was reduced due to elective drop-out, which is characteristic of longitudinal studies. Our findings demonstrate the non-random loss of participants from this cohort and highlight problems that may arise when such data are used in genetic association studies.