The present paper shows that the variability in the cyclic fatigue crack growth (FCG) rate in structural adhesives that is frequently observed in laboratory test results can be captured by using a Hartman-Schijve methodology. To this end the Hartman-Schijve equation has been used to access an ‘upper-bound’ FCG rate curve that (a) encompasses all the experimental data, (b) provides a conservative, ‘worst-case’ FCG rate curve, and (c) accounts for the experimental scatter that is frequently seen under fatigue loading. The importance of allowing for the variability in the measured FCG rate is illustrated by considering FCG in a bonded double-overlap joint subjected to an industry-standard, combat-aircraft flight-load fatigue-cycle spectrum (i.e. FALSTAFF).