The motion of single bubbles rising in 2-octanol solutions was investigated experimentally. By using a high-speed video system to follow the rising bubble, the sequences of the recorded frames were digitized and analyzed using image analysis software. The periodical fluctuation of the bubble terminal velocity was observed, which is indicative of a nonconstant bubble drag coefficient. Then, the measured drag coefficient was compared with correlations available in the literature. The comparison shows that these correlations cannot give fully satisfactory results in predicting the fluctuated drag coefficient. Using the extensive experimental data, the authors proposed a new correlation to predict the drag coefficient calculated from the fluctuated motion with the added mass force and history force included. In virtue of nonlinear curve fitting, the drag coefficient of the single bubble is correlated as a function of the Re number, Eo number, We number, and Mo number based on the equivalent bubble diameter. This new model successfully predicts the periodical fluctuations of the drag coefficients. For the standard drag prediction (when averaged terminal velocity is used), this model agrees better with our experimental results and has a maximum relative error of 2.2% and an average relative error of 0.87%. This model is also verified by bubble rising in methyl isobutyl carbinol cas (MIBC) and OP-10 (C 32 H 58 O 10 ) aqueous solutions, and the errors are similar to that of 2-octanol solutions. On the basis of a comparison with measurements in other liquids from the literature, this new correlation represents well the experimental data which covers 21 pure liquids and contaminated liquids in a very wide range of parameters as follows: 10 −3 ≤ Re ≤ 10 5 , 10 −2 ≤ Eo ≤ 10 3 , and 10 −14 ≤ Mo ≤ 10 7 . This work is an extension of our previous one (Yan et al. Chem. Eng. J. 2017, 316, 553) in which the old correlation predicts the fluctuated drag coefficient for bubble rising in pure water.