Objective To document the influences of blood sampling under supine fasting versus seated nonfasting conditions on diagnosis of phaeochromocytomas and paragangliomas (PPGL) using plasma concentrations of normetanephrine, metanephrine and methoxytyramine. Design and methods Biochemical testing for PPGL was performed on 762 patients at six centres, two of which complied with requirements for supine sampling after an overnight fast and four of which did not. Phaeochromocytomas and paragangliomas were found in 129 patients (67 noncompliant, 62 compliant) and not in 633 patients (195 noncompliant, 438 compliant). Results Plasma concentrations of normetanephrine and methoxytyramine did not differ between compliant and noncompliant sampling conditions in patients with PPGL but were 49-51% higher in patients without PPGL sampled under noncompliant compared with compliant conditions. The 97·5 percentiles of distributions were also higher under noncompliant compared with compliant conditions for normetanephrine (1·29 vs 0·79 nmol/l), metanephrine (0·49 vs 0·41 nmol/l) and methoxytyramine (0·42 vs 0·18 nmol/l). Use of upper cut-offs established from seated nonfasting sampling conditions resulted in substantially decreased diagnostic sensitivity (98% vs 85%). In contrast, use of upper cut-offs established from supine fasting conditions resulted in decreased diagnostic specificity for testing under noncompliant compared with compliant conditions (71% vs 95%). Conclusions High diagnostic sensitivity of plasma normetanephrine, metanephrine and methoxytyramine for the detection of PPGL can only be guaranteed using upper cut-offs of reference intervals established with blood sampling under supine fasting conditions. With such cut-offs, sampling under seated nonfasting conditions can lead to a 5·7-fold increase in false-positive results necessitating repeat sampling under supine fasting conditions.