Vehicle crashes account for the highest number of fatalities for persons aged between 17 and 25 years of age in New Zealand. Despite a myriad of factors precipitating vehicle crashes, excess or inappropriate vehicle speed has been identified as the greatest predictor of crash likelihood and severity. Excess or inappropriate speed reduces a driver’s control over the vehicle, while exaggerating both collision force and the distances required in stopping or safely manoeuvring. One of the major differences identified between young and inexperienced and older more experienced drivers is the ability to adapt driving behaviour to road conditions. Young drivers are more prone to speeding through both a lack of awareness of risks and a desire to seek out novel and stimulating experiences. Recent developments in cognitive models of risk taking propose that older more experienced drivers may adapt their speed by “feeling out” the road conditions, where as young drivers may depend more upon posted limits to determine their speed. A video speed task was developed to measure speed preferences on a selection of road conditions (or ‘environments’) commonly confronting New Zealand motorists. Analyses of speed preferences revealed that young and inexperienced drivers preferred speeds close to the road-limit irrespective of conditions, whereas older and more experienced drivers preferred speeds clearly below the road limit, and demonstrated greater variation in speed preferences on different road environments. This suggests that young and inexperienced drivers both prefer faster speeds and may use the road limit as a target in determining an appropriate speed. Older and more experienced drivers prefer slower speeds, and adapt driving to changing road conditions. Faster preferred speeds were found to be related to a riskier attitudes towards driving in general, and more lenient attitudes toward speeding in particular. In addition, faster preferred speeds were found to be related to a heightened enjoyment of risk taking, as well as the number of speeding convictions issued in the previous 12 months. The used video speed task provided a convenient measure of speeding behaviour in natural driving scenarios, and appeared to be sensitive to differences in the way drivers adjust their behaviour across changing driving conditions. The video speed task might be useful in determining differences in speed choice between day and night time driving scenarios, as well as expanding the road conditions to including wet or foggy driving situations. This may be particularly useful in determining the pre- and post-effectiveness of driver training programs.