The aim of this study was to verify the efficacy and safety of an herbal medication containing an extract of a mixture of Baptisiae tinctoriae radix, Echinaceae pallidae/purpureae radix and Thujae, occidentalis herba (SB-TOX) in the treatment of upper respiratory tract infections (URIs), and to test whether SB-TOX's clinical efficacy is dose dependent. A total of 91 adults (mean age 42.1 +/- 13.0 years) were randomised to receive 19.2 mg of SB-TOX (n=31), 9.6 mg SB-TOX (n=29) or placebo (n=31) three times daily for 3-12 days. Since a "running nose" is the main symptom of a common cold, the total number of facial tissues used throughout the clinical duration of their cold was the primary efficacy parameter. In the intention-to-treat analysis, this total number of tissues decreased with increasing extract dose. The slope across groups according to the Jonckheere test was significant (p = 0.0259). In the high-dose group, the standardised effect size delta/SD was 0.46 compared with placebo. Time to relevant improvement in cold symptoms (measured as the time until less than 30 tissues per day were used) was 1.1 days (95% CI 0.52; 1.67), 0.76 days (95% CI 0.28; 1.24) and 0.52 days (95% CI 0.22; 0.82) in the placebo, low-dose and high-dose groups, respectively (p(LogRank) = 0.0175). No adverse events were reported. This study demonstrates the efficacy and safety of SB-TOX in the treatment of URIs, and that its efficacy is dose dependent.