Sophisticated pacemakers now afford greater benefit than only relief of syncope for patients with symptomatic bradyarrhythmia. For clinical reasons it is inappropriate to implant ventricular demand (VVI) pacemakers in every patient, but for both clinical and economic reasons it is also undesirable to use dual-lead or rate-responsive systems indiscriminately. However, the published guidelines for pacemaker choice and investigation are inadequate. We have analysed the clinical assessment and the investigations required in a series of 150 patients with various electrophysiological disorders, considered consecutively for pacemaker implantation using an algorithm to assist decisions. Holter electrocardiography in 77.3%, exercise testing--formal (in 9.3%) or informal (during Holter electrocardiography), electrophysiology study (EPS)--full in 15.3% and partial in 29.3%, and trial of pacing (10.7%) were all used to characterize dysfunction and to define appropriate pacing mode. 142 patients were paced--57.0% with single lead modes (12.7%) AAI, 44.3% VVI) and 43.0% dual-lead (2.8% DVI, 23.9% VDD or VAT, 16.2% DDD). Three patients received rate-responsive systems (VVIR) which, if available, would have been used in nine; four would have been suitable for AAIR and four for dual-lead responsive pacing (DVIR, DDIR or DDDR). We offer a rational approach based on published data for investigation and pacemaker choice in the management of patients with bradyarrhythmia.