It is unclear if the dye-dilution theory and its corresponding parameters are capable of measuring brain perfusion using harmonic grey-scale imaging. We performed a study on a flow phantom using a SONOS 5500 (1.8--3.6-MHz harmonic imaging) and Levovist as the ultrasound (US) contrast agent (UCA). We applied the UCA in six different doses (0.1 to 3.0 mL) and used eight different flow-rates (180 to 540 mL/min). Additionally, we performed a study on dog brain using Levovist boluses of 1.5 mL and 3 mL. We evaluated the influence of dose and flow-rate on the parameters of the time-intensity curve: peak signal intensity (PSI), area under the curve (AUC) and mean transit time (MTT). Along with an increase of the Levovist dose, the AUC and the PSI increased only in the dose range between 0.1 and 0.5 mL Levovist; further increase led to no change of parameters. Flow-rate showed no influence on AUC, MTT or PSI. The dye-dilution theory is not a useful theoretical model for the analysis of perfusion using harmonic grey-scale imaging. A possible explanation for this effect is the bubble saturation.