OBJECTIVE: This study examined patterns, determinants, and costs of seeking care for mild to moderate psychiatric distress in order to determine optimal approaches for expanding mental health care in rural Haiti. METHODS: A cross-sectional, zone-stratified household survey of 408 adults was conducted in Haiti's Central Plateau. Multivariable logistic regression models were built to assess determinants of first-choice and lifetime health service use by provider type. RESULTS: Thirty-two percent of respondents endorsed God as their first choice for care if suffering from mental distress, and 29% of respondents endorsed clinics and hospitals as their first choice. Forty-seven percent of respondents chose potential providers on the basis of anticipated efficacy. Suicidal individuals were 7.6 times (95% confidence interval [CI]=1.4-42.0) as likely to prefer community-based providers (herbal healer, church priest or pastor, or Vodou priest) over hospitals or clinics. Depression severity was associated with increased odds (adjusted odds ratio [AOR]=1.8, CI=1.5-2.3) of ever having been to an herbal healer. Having a household member with mental health problems was associated with increased odds of ever having been to church pastors or priests (AOR=5.8, CI=2.8-12.0) and decreased odds of ever having been to hospitals or clinics (AOR=.3, CI=.1-.8). Median actual service costs were US $1 for hospitals or clinics, $6 for herbal healers, and $120 for Vodou priests. CONCLUSIONS: Three out of four rural Haitians said they would seek community resources over clinical care if suffering from mental distress. Therefore, isolated clinical interventions may have limited impact because of less frequent use. Efforts to expand mental health care should consider differential provider costs when selecting community resources for task shifting.