Queue formation and dissipation have been extensively studied in relation to traffic signalization, work zone operation, incident occurrence, and ramp metering. This study is an attempt to estimate the effect of vehicle mix, commute time, traffic direction, and road upgrade on queue dissipation speed (time). The data were collected at several intersections in Davis, California, U.S. and analyzed using regression models. The models were determined regressing several functional forms and considering the statistical significance and ease of interpretation of the included variables. The main findings are: 1) dissipation speed does not vary purely by location; 2) a heavy vehicle is faster to discharge than its passenger car sizeequivalent is; 3) the queue in a left-turn lane discharges faster than that in a through lane; 4) an upgrade slope increases the queue dissipation time due to more rolling resistance to vehicle start-up and larger vehicle gaps for safety ; 5) morning queues generally discharge more slowly; 6) contrary to common delay estimation models, regression analysis shows that queue dissipation time is linearly related to the number of vehicles rather than quadratically or in other ways; and 7) the simple linear function performs well both in terms of explanatory power (R2) and consistency of signs.