OBJECTIVES:  To quantify and cross-validate resource-use and cost estimates for a patient cohort with schizophrenia using two data collection methods. METHODS:  Adult patients with persistent symptoms of schizophrenia despite adequately dosed antipsychotic treatment without an acute exacerbation in 3 months were recruited. Resource-use data for healthcare services (inpatient, outpatient, community) and health professional contacts were collected using Electronic Service Use (ESU; 1 year) data and the self-reported Client Service Receipt Inventory (CSRI; 3 months). Unit costs from national and published sources were used to produce cost estimates from an English NHS perspective. Agreement between data collection methods was assessed using Lin’s Concordance Coefficient (pc) and Cohen’s cut-offs for describing strength of agreement (pc: <0.3, weak; 0.3<0.5, moderate; ≥0.5, strong). RESULTS:  For 304 patients over a 3 month period, ESU or CSRI data reported the mean per patient healthcare services cost (contacts) as £548 (3.08) or £1,368 (7.13), respectively. When accounting for other health professional contacts this mean cost increased to £627 or £1,688, respectively. Lin’s coefficient implied strong agreement between data collection methods for healthcare service cost estimates (pc = 0.510), although agreement was weak for resource-use estimates (pc = 0.157). When extrapolating the results to an assumed English patient population with schizophrenia (approximately 266,650 patients), annual healthcare service costs were estimated to between £1.1 and £2.7 billion – dependent on method of data collection. CONCLUSIONS: The two data collection methods produced different results, although strength of agreement between cost estimates was described as strong when using lenient cut-offs. The ‘real’ effect was that self-reported data produced cost-estimates 2.5 times higher than the electronic data. However, even when using conservative estimates, direct healthcare service costs for schizophrenia are high. This study adds detailed resource-use and cost estimates to the current literature on the economic burden of schizophrenia to the English NHS.