Subclinical hypothyroidism (SCH) is a common condition that is often observed without therapy. However, no evidence-based recommendation exists with regards to how patients with untreated SCH should be monitored. Monitoring involves regular assessment of symptoms and signs of hypothyroidism (HYPO) and biochemical tests of thyroid function. An important question when repeated tests of thyroid function are performed is how large a difference in test results is needed to be confident that the change is real and not just due to chance variation. Recent data show that the least significant difference between two tests in SCH is 40% for TSH and 15% for free thyroxine and free triiodothyronine, with 90% confidence. Furthermore, monitoring has to be based on biochemical function testing because serial evaluation of symptoms and signs related to HYPO is rather insensitive in detecting worsening of thyroid insufficiency. When the presence of thyroid peroxidase auto-antibodies (TPO-Ab) in serum has been demonstrated, repeated measurements do not add much useful information in the monitoring of individual subclinical hypothyroid patients, as levels of TPO-Ab vary in parallel with TSH in these patients. Lastly, we discuss how differences in the monitoring procedure influence the diagnostic outcome, and we suggest a follow-up approach for untreated subclinical hypothyroid patients.