Does a fun-riding driving attitude—as a driving safety attitude that is closely related to risky driving—affect a driver's lane-changing decision, and if so, under what circumstances? This study aimed to examine the interactions between road situations, fun-riding driving attitude, areas of interest in lane-changing decisions, and the allocations of visual attention of drivers. Road situations, fun-riding driving attitude, and areas of interest were independent variables in this study. Road situations and areas of interest were the within-group variables. The tendency of fun-riding driving attitude was the between-groups variable. Reserved lane-changing time and total fixation durations on the areas of interest were dependent variables. Using the position of the lane-changing vehicle on the lane as the dividing standard, road situations in this study were classified into two types: one-sided and two-sided lane-change situations. In the one-sided lane-change situations, the experimental vehicles were driving on the inner or outer lane. Only one lane can be targeted for changing lanes. But in two-sided lane-change situations, the experimental vehicles were driving in the middle lane. Both the left and right lanes could be used as target lanes. The revised Chinese version of the Driving Safety Attitude Scale (DSAS-C) was used to divide participants into higher and lower fun-riding driving attitude tendency groups. Thirty-five drivers were recruited for experiments in Dalian, China. They ranged in age from 19 to 50 years old and the average age was 28 years old. Participants were shown videos taken from the driver's perspective through the computer and were asked to change lanes by clicking the left mouse button. We employed the Tobii eye tracker to collect the eye movement markers and reaction time of participants. The sampling rate of eye movement data was over 75% for each of them. The eye movement heatmap of drivers was used to classify the areas of interest in the front view and side mirrors. An analysis of variance showed that fun-riding driving attitude tendency was a predictor of the time reserved for lane-changing decisions in two-sided lane-change situations. Drivers with a higher tendency of fun-riding driving attitude reserved a shorter time for changing lanes in the situations. We also found that tendency of fun-riding driving attitude and areas of interest interacted with the total fixation durations of the drivers. Compared to the drivers with a lower tendency of fun-riding driving attitude, drivers with a higher tendency were more focused on the front view in two-sided lane-change situations. Moreover, they also paid more attention to the side mirrors in one-sided lane-change situations. A path analysis of the structural equation model showed that total fixation duration on the front view could be regarded as a mediator of the influence of fun-riding driving attitude tendency on reserved lane-changing time in two-sided lane-change situations. The results showed that the higher the driver's fun-riding driving attitude tendency was, the more he would focus on the front view and the more impulsive the lane-change decision would be in two-sided lane-change situations. The results support the application of the theory of planned behavior in the study of driving decision-making. They are believed to provide evidence for the effectiveness of eye movement measurement methods in measuring and predicting behavioral intentions.