The purpose of this study was to assess the effects of backward cycling on posterior protective stepping responses in people with Parkinson disease. Secondary aims were to assess changes in gait parameters, balance, and aerobic capacity and to assess the acceptability of backwards cycling as an intervention. Twenty-two people (18 males, 4 females) with a history of idiopathic Parkinson disease in Hoehn & Yahr stages II or III completed a 6-week backward cycling program. Each participant cycled for 30 minutes at a moderate intensity twice a week for 12 total sessions. Pre-test, post-test, and 1-month follow up assessments were completed. Data collected included the Mini-BESTest, 6 Minute Walk Test, posterior stepping response variables (number of steps, time to steady, and maximal excursion), and gait parameters in both the forward and backward directions (gait velocity, right and left step length, and right and left step width). Statistical analysis was performed to assess difference across time and within groups. A repeated measures ANOVA with Bonferroni post hoc analysis was used for normally distributed data. A Friedman ANOVA with a Wilcoxon signed-ranks post hoc test was used for non-normally distributed data. Significant within-group differences were shown in forward and backward gait velocity (p = .017, p = .001), forward right and left step length (p = .011, p = .007), number of steps during a posterior stepping response (p = .019), and the Mini-BESTest (p = .003). Pair-wise post hoc analysis revealed differences between baseline measurements and 1-month follow up in forward right and left step length (p = .049, p = .039), backward gait velocity (p = .007), number of steps during a posterior stepping response (p = .013), and the Mini-BESTest (p < .001). These results show that a backward cycling program is feasible and can have a positive effect on posterior protective stepping responses as well as quality of gait and balance.