In large community surveys, there is a consistent absolute excess of psychological caseness in subjects on thyroxine (T4) compared to age and sex matched controls of around 6%. Part of this likely to be confounding by indication i.e. that some individuals failed to respond to T4 as hypothyroidism was not the cause of their psychological morbidity in the first place. However, the observations that normalisation of intracellular thyroid hormone levels is not achievable with T4 alone in animal models, the fall in T3 levels that occurs after beginning T4, and the preliminary observation subjects with a common polymorphism of deiodinase 2 have worse psychological well being on T4 and appear to benefit from T3, suggest that there may be a subsection of individuals who are not returned o health by T4 alone. Although many randomised studies show no benefit from T4 vs combination studies, studies using higher T3:T4 ratios appear to show some benefit. Despite this, thyroid hormone over-replacement is likely to increase the risk of atrial fibrillation and osteoporosis. Taken together, this information suggests that subjects who report a profound effect on their quality of their life on T4 and are aware of the evidence of potential long-term side-effects of over re-placement should be considered for a trial of T3/T4 until further research information becomes available if they request it. Combination therapy should not be offered routinely, and every effort should be made to minimise doses, emphasise, and mitigate the potential risks and to stop therapy if benefit is not apparent.