Practical challenges related to accurate quantification of carnosic acid (CA), carnosol (CAR) and other phenolic diterpenes in extracts of rosemary leaves (Rosmarinus officinalis L.) are presented and discussed. Primary standard material of CA is isolated from rosemary extracts by preparative chromatography with a purity of 98% or higher. The response factors of CAR relative to CA, at 230 and 280 nm, have been estimated to be 0.92 and 1.36, respectively. The stability of pure CA and CAR, dissolved in methanol, dimethyl sulfoxide (DMSO), DMSO-acetonitrile (10:90) or ethyl acetate-acetonitrile (10:90) and stored at room temperature in the autosampler, is presented. Pure CA dissolved in DMSO is stable for several days, while CAR showed significant degradation within a few hours in all solvents tested. The lack of stability of standards results in practical difficulties with calculating reliable response factors. A correction procedure is presented and documented. A CAR calibration solution was analysed six times for purity during 30 h of storage, while the purity changed from 95 to 70%. Applying this correction procedure resulted in a relative standard deviation on the average response factor of 0.7% (n=6). CA has been dissolved in methanol and stored in clear and amber glass vials, respectively. The solution stored in amber vials degraded faster that in clear vials. The high content of Ti and Fe ions in amber glass seems to catalyse the degradation of CA. In contrast to solutions of pure CAR and CA, their stabilities in solutions of rosemary extracts are fine. A standard addition experiment, covering a time interval of 21 h, resulted in recoveries of CAR and CA of 100 and 96%, respectively.