<b><i>Background:</i></b> While peak in- and expiratory flow rates offer valuable information for diagnosis and monitoring in respiratory disease, these indices are usually considered too variable to be routinely used for quantification in clinical practice. <b><i>Objectives:</i></b> The aim of the study was to obtain reproducible measurements of maximal inspiratory flow rates and to construct reference equations for peak in- and expiratory flows (PIF and PEF). <b><i>Method:</i></b> With coaching for maximal effort, 187 healthy Caucasian subjects (20–80 years) performed at least 3 combined forced inspiratory and expiratory manoeuvres, until at least 2 peak inspiratory flow measurements were within 10% of each other. The effect on PIF preceded by a slow expiration instead of a forced expiration and PIF repeatability over 3 different days was also investigated in subgroups. Reference values and limits of normal for PIF, mid-inspiratory flow, and PEF were obtained according to the Lambda-Mu-Sigma statistical method. <b><i>Results:</i></b> A valid PIF could be obtained within 3.3 ± 0.6(SD) attempts, resulting in an overall within-test PIF variability of 4.6 ± 3.2(SD)%. A slow instead of a forced expiration prior to forced inspiration resulted in a significant (<i>p</i> &#x3c; 0.001) but small PIF increase (2.5% on average). Intraclass correlation coefficient for between-day PIF was 0.981 (95% CI: 0.960–0.992). Over the entire age range, inter-subject PIF variability was smaller than in previous reports, and PIF could be predicted based on its determinants gender, age, and height (<i>r</i>² = 0.53). <b><i>Conclusions:</i></b> When adhering to similar criteria for the measurement of effort-dependent portions of inspiratory and expiratory flow-volume curves, performed according to current ATS/ERS standards, it is possible to obtain reproducible PIF and PEF values for use in routine clinical practice.