The efficacy of an Actinobacillus pleuropneumoniae subunit vaccine based on ApxIA, ApxIIA, ApxIIIA and OMP-2 (Porcilis App, MSD) was investigated in two farrow-to-finish pig herds (A and B) affected by chronic pleurisy. In total, 1161 pigs were included. At three weeks of age, the pigs were randomly allocated to non-vaccinated control (NV; n=580) and vaccinated (V; n=581) groups. At 6 and 10 weeks of age, pigs were injected with Porcilis-APP (V group) or adjuvant (NV group). At slaughter (26 weeks), pleurisy and pneumonia lesions were assessed. All pigs were weighed individually at 6 and 26 weeks of age, and average daily weight gain (ADG; g/pig/day) was calculated. Mortality and days of additional treatment (DAT) were registered during the whole experiment. Data were analysed using binary logistic regression or analysis of variance for proportions or continuous variables, respectively. The prevalence of pleurisy and pneumonia was (NV-A=19.3, V-A=7.9, (P=0.000); NV-B=17.9, V-B=0.7, (P=0.000)) and (NV-A=42.4, V-A=21.2, (P=0.000); NV-B=46.7, V-B=19.0, (P=0.000)), respectively. The ADG was NV-A=632±157, V-A=647±91, (P=0.162); NV-B=660±115, V-B=670±82, (P=0.232). The mortality during the experiment was NV-A=5.7, V-A=1.8, (P=0.015); NV-B=2.3, V-B=1.0, (P=0.170) per cent. The DAT was: NV-A=15.04±1.41, V-A=14.95±0.67, (P=0.010); NV-B=21.68±2.43, V-B=16.99±0.62, (P=0.000). The present study showed a significant reduction of the prevalence of pleurisy and pneumonia, and antimicrobial use in V pigs from both herds, and in mortality in V pigs from one herd.