Background <p>Advanced platelet-rich fibrin (A-PRF) is used in periodontal regeneration due to its autologous fibrin matrix and bioactive content. Collagen membranes are widely used in guided tissue regeneration. However, it remains unclear whether combining A-PRF with a collagen membrane alters early cellular responses. This in vitro study compared the effects of A-PRF alone and A-PRF combined with a collagen membrane on the viability, proliferation-related metabolic activity, and migration-related wound closure of human periodontal ligament stem cells (hPDLSCs) at two extract concentrations.</p> Methods <p>hPDLSCs were exposed to extracts prepared from A-PRF alone or A-PRF combined with a porcine collagen membrane (A-PRF-Col) at 100% and 20% concentrations. Extracts were prepared using A-PRF derived from the blood of four donors. Cell viability was assessed after 24&#xa0;h using an MTT assay and morphological observation. Proliferation-related metabolic activity was evaluated by MTT assay on days 1, 3, 5, and 7. Migration-related wound closure was evaluated using a scratch assay after 24&#xa0;h. Data were analysed using repeated-measures ANOVA with post hoc multiple-comparison tests. <i>p</i> &lt; 0.05 was considered statistically significant.</p> Results <p>Early hPDLSC responses differed between A-PRF and A-PRF-Col and varied between the two extract concentrations. Based on mean relative growth rate values, only 100% A-PRF-Col met the predefined cytotoxicity criterion, whereas 100% A-PRF showed a borderline and donor-variable viability profile. Proliferation-related metabolic activity increased and peaked around day 5, but different concentrations and formulations affected the patterns. Migration-related wound closure was primarily influenced by extract concentration, with 20% extracts showing greater closure than their 100% counterparts. At 20% concentration, A-PRF achieved better closure than A-PRF-Col.</p> Conclusions <p>Under the present cell-source and extract-preparation conditions, hPDLSC responses differed between A-PRF and A-PRF-Col and varied between the two tested extract concentrations. Diluted extracts generally elicited more favourable early cellular responses than undiluted extracts. The A-PRF-Col construct did not uniformly enhance A-PRF effects across endpoints. These findings should be interpreted as exploratory evidence that A-PRF-Col has an endpoint- and concentration-related biological profile that requires further optimisation.</p> Trial registration <p>Not applicable.</p>