OBJECTIVES: To describe incremental direct and indirect costs among migraine patients compared to matched controls in the United States. METHODS:  This retrospective cohort study identified migraine patients (ICD-9 code 346.XX and/or migraine-specific medications) in the Truven Health MarketScan Commercial and Health and Productivity Management Databases between January 2008 and June 2013. Adult patients had 12 months of continuous enrollment before and after the day they received migraine diagnoses and/or medications (index) and no diagnosis of HIV or malignancy during the study period. A group of non-migraine patients was matched 1:1 on age, gender, region, health plan type and index date (randomly assigned). Direct (inpatient stays, outpatient visits, emergency room visits, and acute and prophylactic migraine medication use) and indirect (absenteeism, short term disability and long term disability) costs were assessed during the 12 month post-index period and compared using chi-square and t-tests. RESULTS:  A total of 84,245 migraine patients met the inclusion criteria. Compared to non-migraine matched controls, migraine patients had significantly higher direct costs, overall ($10,363 [SD=$17,058] vs. $4,619 [SD=$10,394]) and in each cost category (emergency room, inpatient, outpatient, pharmacy) (all p<0.001). Migraine patients had higher absenteeism costs than controls ($9,304 vs. $8,261), driven by more absenteeism days during follow-up (33.2[SD=21.0) vs. 30.7[SD=16.4], p<0.001). Migraine patients were 2.5 times as likely to have a short term disability claim than matched controls (16.7% vs. 6.7%, p<0.001), with higher costs for short term disability ($1,828 vs. $603). Migraine patients were also 2.4 times more likely (0.61% vs. 0.25%) to have a long term disability claim, with an average cost of $26,543 per claim. Overall, migraine patients had higher indirect costs compared to non-migraine controls ($11,294 vs. $8,945). CONCLUSIONS:  Migraine patients incur considerable direct and indirect costs which is significantly higher than matched non-migraine patients.