Threshold values for long and short cracks play a key role in damage tolerant design. However, these values depend very much on the experimental method used. The classical procedure for evaluating the threshold value for long cracks is the load shedding method where the range of the stress intensity factor is gradually reduced until the crack velocity drops below a pre-defined value. It has been pointed out that the crack tip is then contained in the plastic zone of previous load cycles and, consequently, is arrested in a zone of hardened material with different material properties than the material of interest. An alternative may be to apply very small amplitudes as the ones used in the VHCF regime and use continuous crack monitoring for measuring the crack advance at very low propagation rates. This idea makes only sense as long as the plastic zone of crack tip field is not changed by the low load level. An experimental investigation was started with the goal of defining threshold values which are independent of prior plastic deformation and can also be used in the VHCF regime. For this purpose sharp micro-notches were introduced in a precipitation-hardened aluminum alloy EN AW 6082 by a special “low-force” saw-cut to avoid prior plastic deformation. Cyclic loading with very low amplitudes was used to advance the crack. The crack growth rate was evaluated from the crack path using a travelling long-distance optical microscope. A crack growth rate of 10-11m/cycle was set as a critical value for the threshold. The crack tip field of the growing crack is evaluated using digital image correlation, EBSD and a stress analysis based on the CrossCourt tool, and crack tip fields of advancing and non-advancing cracks are compared. Moreover, the effect of the microstructure is taken into consideration by considering crack advancing along the rolling direction and perpendicular to it.