BACKGROUND To reduce travel burdens, patients may coordinate outpatient services to receive multiple encounters during a single day. Unmeasured visit "chaining" may bias estimates of the impact of accessibility barriers when utilization volume is measured using visit days. No studies have evaluated differential encounter chaining by distance. OBJECTIVE We sought to evaluate the prevalence and predictors of visit chaining among patients with serious mental illnesses (SMIs), to evaluate whether patients living farther from providers are more likely to chain encounters, and to assess distance barriers using alternative measures of utilization volume. METHODS We used the Veterans Affairs (VA) National Psychosis Registry, including Fiscal Year 2000 diagnosis and utilization data for 141,275 VA patients with SMI diagnoses. METHODS Random intercepts hierarchical regression to examine the relationship between distance to nearest VA provider and encounters/visit day, adjusting for age, gender, race/ethnicity, marital status, service-connection, previous hospitalization, and treatment setting and comorbidities at initial use. Negative binomial regressions to evaluate distance effects on visit day and encounter volume. Analyses adjusted for patient clustering within facilities. RESULTS With increased distance, patients had more encounters/visit day. Patients with bipolar disorder were more likely than patients with schizophrenia or other psychoses to chain treatments in association with greater distances. CONCLUSIONS When utilization volume is measured in terms of visit days, analyses may overestimate distance barriers, because remote patients are more likely to chain encounters within visit days. However, distance remains a substantial barrier limiting total outpatient visit volume. Enhanced services coordination may reduce accessibility barriers for remote patients.