Transportation demand management policies are increasingly deployed to mitigate urban congestion, yet their effectiveness may vary systematically across trip purposes—a dimension often overlooked in policy evaluation. This study investigates how work and non-work travelers respond differently to parking pricing and bus travel time improvements in the central business district of Isfahan, Iran. Using stated preference data from 428 private vehicle users, separate binary and multinomial logit models are estimated for work and non-work trips. Results reveal substantial heterogeneity: work trips exhibit lower parking price elasticity (-0.35) than non-work trips (-0.42), reflecting greater schedule rigidity and reliability requirements of commuting. Bus travel time improvements are more effective for non-work trips, while park-and-ride facilities attract a disproportionately higher share of work travelers. Socio-demographic moderators also differ: women and unmarried individuals show higher car preference for work trips, whereas household vehicle ownership and high-value vehicles drive car choice for non-work trips. These findings suggest that uniform travel demand management policies may be suboptimal; instead, trip-purpose-specific policy packages—combining pricing with reliability-enhancing transit for commuters and pricing with frequency improvements for discretionary travelers—can achieve greater effectiveness. The study contributes empirical evidence on behavioral heterogeneity from a middle-income country context and offers actionable guidance for differentiated congestion management in historic city centers.