The structure of human fibroblasts have been characterised in vitro by atomic force microscopy (AFM) operated in the imaging or in the force versus distance (F‐d) modes. The choice of cell substrate is important to ensure good adhesion. Of greater significance in the context of AFM analysis, is the observation that the substrate affects the imaging conditions for in vitro analysis of live cells. For instance, very rarely will glass coverslips lead to acceptable outcomes (i.e., resolved cytoskeletal structure). Activated tissue culture dishes, on the other hand, promote conditions that routinely result in good quality images. Those conditions are then unaffected by adoption of relatively high force loadings (more than 10 nN), large fields of view (100×100μm2) and high scan speeds (up to ca. 200 μm/sec), all of which exceed values recommended in the literature. Plasma membranes are fragile in the context of AFM analysis (F‐d analysis gives an equivalent Young's Modulus of ca. 5 kPa). However, the present work suggests that fragility per se need not be a problem, rather it is the adhesive interactions with the tip, which under some circumstances may exceed 20 nN, that are the source of poor imaging conditions. The present results, being supported by a qualitative model, suggest that the activated substrate acts as a preferential scavenger of cellular debris thus preventing the tip from biofouling, and will therefore promote low adhesion between tip and membrane. Good imaging conditions provide non‐destructive in vitro information about cytoskeletal structure and dynamics, as shown in two examples concerned with cytochalasin treatment and with the MTT assay.