Some patients do not feel well on L-thyroxine despite a serum TSH in the reference range. Key issues to consider in such patients include establishing whether the patient had established hypothyroidism initially, and whether the Lthyroxine has been titrated carefully enough, possibly using small increments, to achieve a careful balance between symptoms and serum TSH concentrations. Patients should also be considered for other causes of the symptoms which may be wide-ranging. Meta-analyses of several, but small, randomised control trials show no advantage, or disadvantage of liothyronine over L-thyroxine. However, detailed sub-analysis identifies some tantalising results eg on preferential weight loss, patient preference, and possibly genetic markers. Although linked with plausible theoretical explanations, these results may be over-interpreted. The key questions are whether a short-term trial treatment is worthwhile and safe, and whether in the future sub-groups of patients can be identified who may benefit from liothyronine. These questions remain divisive but require additional focussed research.