Needle trap devices (NTDs) are a relatively new and promising tool for headspace (HS) analysis. In this study, a dynamic HS sampling procedure is evaluated for the determination of volatile organic compounds (VOCs) in whole blood samples. A full factorial design was used to evaluate the influence of the number of cycles and incubation time and it is demonstrated that the controlling factor in the process is the number of cycles. A mathematical model can be used to determine the most appropriate number of cycles required to adsorb a prefixed amount of VOCs present in the HS phase whenever quantitative adsorption is reached in each cycle. Matrix effect is of great importance when complex biological samples, such as blood, are analyzed. The evaluation of the salting out effect showed a significant improvement in the volatilization of VOCs to the HS in this type of matrices. Moreover, a 1:4 (blood:water) dilution is required to obtain quantitative recoveries of the target analytes when external calibration is used. The method developed gives detection limits in the 0.020–0.080 μg L−1 range (0.1–0.4 μg L−1 range for undiluted blood samples) with appropriate repeatability values (RSD < 15% at high level and <23% at LOQ level). Figure of merits of the method can be improved by using a smaller phase ratio (i.e., an increase in the blood volume and a decrease in the HS volume), which lead to lower detection limits, better repeatability values and greater sensibility. Twenty-eight blood samples have been evaluated with the proposed method and the results agree with those indicated in other studies. Benzene was the only target compound that gave significant differences between blood levels detected in volunteer non-smokers and smokers