The purpose of this work is to evaluate if the coffee-associated malabsorption of tablet levothyroxine (L-T4) is reduced by soft gel capsule. We recruited 8 patients with coffee-associated L-T4 malabsorption including one hypothyroid patient. For 6 months, the patients were switched to the capsule maintaining the L-T4 daily dose. Patients took the capsule with water, having coffee 1 h later (proper habit, PH) on days 1-90, or with coffee ≤ 5 min later (improper habit, IH) on days 91-180. After 6 months, 2 patients volunteered for an acute loading test of 600 μg L-T4 (capsule) ingested with water (PH) or with coffee (IH). In the single hypothyroid patient, the post-switch TSH ranged 0.06-0.16 mU/L (PH) versus 5.8-22.4 mU/L pre-switch (PH) and 0.025-0.29 mU/L (IH) versus 26-34 mU/L pre-switch (IH). In the other 7 patients, post-switch TSH was 0.41 ± 0.46 (PH) versus 0.28 ± 0.20 pre-switch (PH) (P = 0.61) and 0.34 ± 0.30 (IH) versus 1.23 ± 1.47 pre-switch (IH) (P < 0.001). Importantly, TSH levels in PH versus IH habit did not differ post-switch (P = 0.90), but they did pre-switch (P < 0.0001). The proportions of post-switch TSH levels <0.10 mU/L with PH (33.3 %) or with IH (33.3 %) were borderline significantly greater than the corresponding pre-switch levels with PH (10.3 %) (P = 0.088) or with IH (0 %) (P = 0.0096). In the two volunteers, the L-T4 loading test showed that coffee influenced L-T4 pharmacokinetics minimally. Soft gel capsules can be used in patients who are unable/unwilling to change their IH of taking L-T4.