Purpose The purpose of this paper is to use a heutagogical approach to determine whether students enrolled in blended courses achieve higher grades relative to those enrolled in completely online courses, in addition to identifying demographic predictors of academic success in college courses involving Web-based modes of instruction. Design/methodology/approach Mixed models regression evaluated predictors of grade in terms of age, gender, instruction mode, graduate vs undergraduate status and full-time vs part-time load across 2,174 students (M = 27.6, SD = 9.54 years) enrolled in Web-based courses for a single term at a mid-sized public university in the northeastern USA. Findings In accordance with expectations, a significant main effect indicated higher grades among students enrolled in blended relative to completely online courses. Other predictors of academic success in Web-based courses included older age, female gender, graduate student status and part-time academic load. An interaction between age and gender on grade indicated the difference in performance between men and women diminished among older compared to younger students. Another interaction between age and instruction mode on successful course completion indicated a higher probability of success in blended courses among older students relative to their younger counterparts. Research limitations/implications This study is limited by its cross-sectional design of large scope, which is incapable of addressing differences in online instructional styles and student motivation factors. Originality/value The current study offers newfound evidence that students enrolled in Web-based college courses may benefit from a blended instructional format, a finding that may be particularly evident among older students.