The effectiveness of manipulation versus mobilization for the management of spinal conditions, including cervicogenic headache, is conflicting. However, a pragmatic approach comparing manipulation to mobilization has not been examined in a patient population with cervicogenic headache. To evaluate the effectiveness of manipulation compared to mobilization applied in a pragmatic fashion for patients with cervicogenic headache. Forty-five (26 females) patients with cervicogenic headache (mean age 47.8 ± SD 16.9 years) were randomly assigned to receive either pragmatically selected manipulation or mobilization. Outcomes were measured at baseline, the second visit, discharge, and 1-month follow-up and included the Neck Disability Index (NDI), Numeric Pain Rating Scale (NPRS), the Headache Impact Test (HIT-6), the Global Rating of Change (GRC), the Patient Acceptable Symptoms Scale (PASS). The primary aim (effects of treatment on disability and pain were examined with a mixed-model analysis of variance (ANOVA), with treatment group (manipulation versus mobilization) as the between subjects variable and time (baseline, 48 hours, discharge and follow-up) as the within subjects variable. The interaction for the mixed model ANOVA was not statistically significant for NDI (p = 0.91), NPRS (p = 0.81), or HIT (p = 0.89). There was no significant difference between groups for the GRC or PASS. The results suggest that manipulation has similar effects on disability, pain, GRC, and cervical range of motion as mobilization when applied in a pragmatic fashion for patients with cervicogenic headaches. NCT03919630 [ABSTRACT FROM AUTHOR]