The thermal diffusivity of various types of aluminum has been measured, using a completely noncontact experimental configuration based on infrared photothermal radiometry. Photothermal response transients, conventional frequency scans, and pulse duration- or repetition rate-scanned rate windows have been investigated. It has been shown that the conventional frequency scan is not suitable for measurements of aluminum with a short thermal transport time such as foils, due to an extremely degraded signal-to-noise ratio (SNR). Also, it has been found that the conventional frequency scan method is less sensitive to the actual value of thermal diffusivity than the rate-window scan. The rate-window method, furthermore, gives superior SNR especially for thin metals and yields excellent agreement between the theory and the data. An advantage of the pulse duration-scanned rate window mode is that it does not require knowledge of the instrumental transfer function as an input. The transient response gives the worst SNR but is best for the physical interpretation of the photothermal signals. In addition, it has been shown that the infrared photothermal radiometric transmission mode is less sensitive to surface roughness than the reflection mode and, therefore, is preferable for thermal diffusivity measurements of aluminum and of good thermal conductors, in general.